fraud detection-alternate implementation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "%matplotlib inline\n",
    "sns.set_style(\"whitegrid\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Time</th>\n",
       "      <th>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>...</th>\n",
       "      <th>V21</th>\n",
       "      <th>V22</th>\n",
       "      <th>V23</th>\n",
       "      <th>V24</th>\n",
       "      <th>V25</th>\n",
       "      <th>V26</th>\n",
       "      <th>V27</th>\n",
       "      <th>V28</th>\n",
       "      <th>Amount</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.00</td>\n",
       "      <td>-1.36</td>\n",
       "      <td>-0.07</td>\n",
       "      <td>2.54</td>\n",
       "      <td>1.38</td>\n",
       "      <td>-0.34</td>\n",
       "      <td>0.46</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.36</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.28</td>\n",
       "      <td>-0.11</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.13</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>0.13</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>149.62</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.00</td>\n",
       "      <td>1.19</td>\n",
       "      <td>0.27</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.45</td>\n",
       "      <td>0.06</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>-0.08</td>\n",
       "      <td>0.09</td>\n",
       "      <td>-0.26</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>-0.64</td>\n",
       "      <td>0.10</td>\n",
       "      <td>-0.34</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.13</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>2.69</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.00</td>\n",
       "      <td>-1.36</td>\n",
       "      <td>-1.34</td>\n",
       "      <td>1.77</td>\n",
       "      <td>0.38</td>\n",
       "      <td>-0.50</td>\n",
       "      <td>1.80</td>\n",
       "      <td>0.79</td>\n",
       "      <td>0.25</td>\n",
       "      <td>-1.51</td>\n",
       "      <td>...</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.91</td>\n",
       "      <td>-0.69</td>\n",
       "      <td>-0.33</td>\n",
       "      <td>-0.14</td>\n",
       "      <td>-0.06</td>\n",
       "      <td>-0.06</td>\n",
       "      <td>378.66</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.00</td>\n",
       "      <td>-0.97</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>1.79</td>\n",
       "      <td>-0.86</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1.25</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.38</td>\n",
       "      <td>-1.39</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.11</td>\n",
       "      <td>0.01</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>-1.18</td>\n",
       "      <td>0.65</td>\n",
       "      <td>-0.22</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.06</td>\n",
       "      <td>123.50</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.00</td>\n",
       "      <td>-1.16</td>\n",
       "      <td>0.88</td>\n",
       "      <td>1.55</td>\n",
       "      <td>0.40</td>\n",
       "      <td>-0.41</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.59</td>\n",
       "      <td>-0.27</td>\n",
       "      <td>0.82</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.80</td>\n",
       "      <td>-0.14</td>\n",
       "      <td>0.14</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.22</td>\n",
       "      <td>69.99</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Time    V1    V2   V3    V4    V5    V6    V7    V8    V9  ...   V21   V22  \\\n",
       "0  0.00 -1.36 -0.07 2.54  1.38 -0.34  0.46  0.24  0.10  0.36  ... -0.02  0.28   \n",
       "1  0.00  1.19  0.27 0.17  0.45  0.06 -0.08 -0.08  0.09 -0.26  ... -0.23 -0.64   \n",
       "2  1.00 -1.36 -1.34 1.77  0.38 -0.50  1.80  0.79  0.25 -1.51  ...  0.25  0.77   \n",
       "3  1.00 -0.97 -0.19 1.79 -0.86 -0.01  1.25  0.24  0.38 -1.39  ... -0.11  0.01   \n",
       "4  2.00 -1.16  0.88 1.55  0.40 -0.41  0.10  0.59 -0.27  0.82  ... -0.01  0.80   \n",
       "\n",
       "    V23   V24   V25   V26   V27   V28  Amount  Class  \n",
       "0 -0.11  0.07  0.13 -0.19  0.13 -0.02  149.62      0  \n",
       "1  0.10 -0.34  0.17  0.13 -0.01  0.01    2.69      0  \n",
       "2  0.91 -0.69 -0.33 -0.14 -0.06 -0.06  378.66      0  \n",
       "3 -0.19 -1.18  0.65 -0.22  0.06  0.06  123.50      0  \n",
       "4 -0.14  0.14 -0.21  0.50  0.22  0.22   69.99      0  \n",
       "\n",
       "[5 rows x 31 columns]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(\"/kaggle/input/creditcardfraud/creditcard.csv\")\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Exploratory Data Analysis\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "_kg_hide-output": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 284807 entries, 0 to 284806\n",
      "Data columns (total 31 columns):\n",
      "Time      284807 non-null float64\n",
      "V1        284807 non-null float64\n",
      "V2        284807 non-null float64\n",
      "V3        284807 non-null float64\n",
      "V4        284807 non-null float64\n",
      "V5        284807 non-null float64\n",
      "V6        284807 non-null float64\n",
      "V7        284807 non-null float64\n",
      "V8        284807 non-null float64\n",
      "V9        284807 non-null float64\n",
      "V10       284807 non-null float64\n",
      "V11       284807 non-null float64\n",
      "V12       284807 non-null float64\n",
      "V13       284807 non-null float64\n",
      "V14       284807 non-null float64\n",
      "V15       284807 non-null float64\n",
      "V16       284807 non-null float64\n",
      "V17       284807 non-null float64\n",
      "V18       284807 non-null float64\n",
      "V19       284807 non-null float64\n",
      "V20       284807 non-null float64\n",
      "V21       284807 non-null float64\n",
      "V22       284807 non-null float64\n",
      "V23       284807 non-null float64\n",
      "V24       284807 non-null float64\n",
      "V25       284807 non-null float64\n",
      "V26       284807 non-null float64\n",
      "V27       284807 non-null float64\n",
      "V28       284807 non-null float64\n",
      "Amount    284807 non-null float64\n",
      "Class     284807 non-null int64\n",
      "dtypes: float64(30), int64(1)\n",
      "memory usage: 67.4 MB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Time</th>\n",
       "      <th>V1</th>\n",
       "      <th>V2</th>\n",
       "      <th>V3</th>\n",
       "      <th>V4</th>\n",
       "      <th>V5</th>\n",
       "      <th>V6</th>\n",
       "      <th>V7</th>\n",
       "      <th>V8</th>\n",
       "      <th>V9</th>\n",
       "      <th>...</th>\n",
       "      <th>V21</th>\n",
       "      <th>V22</th>\n",
       "      <th>V23</th>\n",
       "      <th>V24</th>\n",
       "      <th>V25</th>\n",
       "      <th>V26</th>\n",
       "      <th>V27</th>\n",
       "      <th>V28</th>\n",
       "      <th>Amount</th>\n",
       "      <th>Class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>...</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "      <td>284807.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>94813.86</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>-0.00</td>\n",
       "      <td>88.35</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>47488.15</td>\n",
       "      <td>1.96</td>\n",
       "      <td>1.65</td>\n",
       "      <td>1.52</td>\n",
       "      <td>1.42</td>\n",
       "      <td>1.38</td>\n",
       "      <td>1.33</td>\n",
       "      <td>1.24</td>\n",
       "      <td>1.19</td>\n",
       "      <td>1.10</td>\n",
       "      <td>...</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.52</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.33</td>\n",
       "      <td>250.12</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.00</td>\n",
       "      <td>-56.41</td>\n",
       "      <td>-72.72</td>\n",
       "      <td>-48.33</td>\n",
       "      <td>-5.68</td>\n",
       "      <td>-113.74</td>\n",
       "      <td>-26.16</td>\n",
       "      <td>-43.56</td>\n",
       "      <td>-73.22</td>\n",
       "      <td>-13.43</td>\n",
       "      <td>...</td>\n",
       "      <td>-34.83</td>\n",
       "      <td>-10.93</td>\n",
       "      <td>-44.81</td>\n",
       "      <td>-2.84</td>\n",
       "      <td>-10.30</td>\n",
       "      <td>-2.60</td>\n",
       "      <td>-22.57</td>\n",
       "      <td>-15.43</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>54201.50</td>\n",
       "      <td>-0.92</td>\n",
       "      <td>-0.60</td>\n",
       "      <td>-0.89</td>\n",
       "      <td>-0.85</td>\n",
       "      <td>-0.69</td>\n",
       "      <td>-0.77</td>\n",
       "      <td>-0.55</td>\n",
       "      <td>-0.21</td>\n",
       "      <td>-0.64</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>-0.54</td>\n",
       "      <td>-0.16</td>\n",
       "      <td>-0.35</td>\n",
       "      <td>-0.32</td>\n",
       "      <td>-0.33</td>\n",
       "      <td>-0.07</td>\n",
       "      <td>-0.05</td>\n",
       "      <td>5.60</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>84692.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.18</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>-0.05</td>\n",
       "      <td>-0.27</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.02</td>\n",
       "      <td>-0.05</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.02</td>\n",
       "      <td>-0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>22.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>139320.50</td>\n",
       "      <td>1.32</td>\n",
       "      <td>0.80</td>\n",
       "      <td>1.03</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.57</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.60</td>\n",
       "      <td>...</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.35</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.09</td>\n",
       "      <td>0.08</td>\n",
       "      <td>77.16</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>172792.00</td>\n",
       "      <td>2.45</td>\n",
       "      <td>22.06</td>\n",
       "      <td>9.38</td>\n",
       "      <td>16.88</td>\n",
       "      <td>34.80</td>\n",
       "      <td>73.30</td>\n",
       "      <td>120.59</td>\n",
       "      <td>20.01</td>\n",
       "      <td>15.59</td>\n",
       "      <td>...</td>\n",
       "      <td>27.20</td>\n",
       "      <td>10.50</td>\n",
       "      <td>22.53</td>\n",
       "      <td>4.58</td>\n",
       "      <td>7.52</td>\n",
       "      <td>3.52</td>\n",
       "      <td>31.61</td>\n",
       "      <td>33.85</td>\n",
       "      <td>25691.16</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Time        V1        V2        V3        V4        V5        V6  \\\n",
       "count 284807.00 284807.00 284807.00 284807.00 284807.00 284807.00 284807.00   \n",
       "mean   94813.86      0.00      0.00     -0.00      0.00     -0.00      0.00   \n",
       "std    47488.15      1.96      1.65      1.52      1.42      1.38      1.33   \n",
       "min        0.00    -56.41    -72.72    -48.33     -5.68   -113.74    -26.16   \n",
       "25%    54201.50     -0.92     -0.60     -0.89     -0.85     -0.69     -0.77   \n",
       "50%    84692.00      0.02      0.07      0.18     -0.02     -0.05     -0.27   \n",
       "75%   139320.50      1.32      0.80      1.03      0.74      0.61      0.40   \n",
       "max   172792.00      2.45     22.06      9.38     16.88     34.80     73.30   \n",
       "\n",
       "             V7        V8        V9  ...       V21       V22       V23  \\\n",
       "count 284807.00 284807.00 284807.00  ... 284807.00 284807.00 284807.00   \n",
       "mean      -0.00     -0.00     -0.00  ...      0.00      0.00      0.00   \n",
       "std        1.24      1.19      1.10  ...      0.73      0.73      0.62   \n",
       "min      -43.56    -73.22    -13.43  ...    -34.83    -10.93    -44.81   \n",
       "25%       -0.55     -0.21     -0.64  ...     -0.23     -0.54     -0.16   \n",
       "50%        0.04      0.02     -0.05  ...     -0.03      0.01     -0.01   \n",
       "75%        0.57      0.33      0.60  ...      0.19      0.53      0.15   \n",
       "max      120.59     20.01     15.59  ...     27.20     10.50     22.53   \n",
       "\n",
       "            V24       V25       V26       V27       V28    Amount     Class  \n",
       "count 284807.00 284807.00 284807.00 284807.00 284807.00 284807.00 284807.00  \n",
       "mean       0.00      0.00      0.00     -0.00     -0.00     88.35      0.00  \n",
       "std        0.61      0.52      0.48      0.40      0.33    250.12      0.04  \n",
       "min       -2.84    -10.30     -2.60    -22.57    -15.43      0.00      0.00  \n",
       "25%       -0.35     -0.32     -0.33     -0.07     -0.05      5.60      0.00  \n",
       "50%        0.04      0.02     -0.05      0.00      0.01     22.00      0.00  \n",
       "75%        0.44      0.35      0.24      0.09      0.08     77.16      0.00  \n",
       "max        4.58      7.52      3.52     31.61     33.85  25691.16      1.00  \n",
       "\n",
       "[8 rows x 31 columns]"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.set_option(\"display.float\", \"{:.2f}\".format)\n",
    "data.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us now check the missing values in the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.isnull().sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Time', 'V1', 'V2', 'V3', 'V4', 'V5', 'V6', 'V7', 'V8', 'V9', 'V10',\n",
       "       'V11', 'V12', 'V13', 'V14', 'V15', 'V16', 'V17', 'V18', 'V19', 'V20',\n",
       "       'V21', 'V22', 'V23', 'V24', 'V25', 'V26', 'V27', 'V28', 'Amount',\n",
       "       'Class'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The only non-transformed variables to work with are:\n",
    "- `Time`\n",
    "- `Amount`\n",
    "- `Class` (1: fraud, 0: not_fraud)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "LABELS = [\"Normal\", \"Fraud\"]\n",
    "\n",
    "count_classes = pd.value_counts(data['Class'], sort = True)\n",
    "count_classes.plot(kind = 'bar', rot=0)\n",
    "plt.title(\"Transaction Class Distribution\")\n",
    "plt.xticks(range(2), LABELS)\n",
    "plt.xlabel(\"Class\")\n",
    "plt.ylabel(\"Frequency\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    284315\n",
       "1       492\n",
       "Name: Class, dtype: int64"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.Class.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Determine the number of fraud and valid transactions in the entire dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of Fraudulant transactions: (492, 31)\n",
      "Shape of Non-Fraudulant transactions: (284315, 31)\n"
     ]
    }
   ],
   "source": [
    "fraud = data[data['Class']==1]\n",
    "normal = data[data['Class']==0]\n",
    "\n",
    "print(f\"Shape of Fraudulant transactions: {fraud.shape}\")\n",
    "print(f\"Shape of Non-Fraudulant transactions: {normal.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How different are the amount of money used in different transaction classes?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Amount</th>\n",
       "      <th>Amount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>492.00</td>\n",
       "      <td>284315.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>122.21</td>\n",
       "      <td>88.29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>256.68</td>\n",
       "      <td>250.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.00</td>\n",
       "      <td>5.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>9.25</td>\n",
       "      <td>22.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>105.89</td>\n",
       "      <td>77.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2125.87</td>\n",
       "      <td>25691.16</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Amount    Amount\n",
       "count  492.00 284315.00\n",
       "mean   122.21     88.29\n",
       "std    256.68    250.11\n",
       "min      0.00      0.00\n",
       "25%      1.00      5.65\n",
       "50%      9.25     22.00\n",
       "75%    105.89     77.05\n",
       "max   2125.87  25691.16"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([fraud.Amount.describe(), normal.Amount.describe()], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Do fraudulent transactions occur more often during certain time frame ?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Time</th>\n",
       "      <th>Time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>492.00</td>\n",
       "      <td>284315.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>80746.81</td>\n",
       "      <td>94838.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>47835.37</td>\n",
       "      <td>47484.02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>406.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>41241.50</td>\n",
       "      <td>54230.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>75568.50</td>\n",
       "      <td>84711.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>128483.00</td>\n",
       "      <td>139333.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>170348.00</td>\n",
       "      <td>172792.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Time      Time\n",
       "count    492.00 284315.00\n",
       "mean   80746.81  94838.20\n",
       "std    47835.37  47484.02\n",
       "min      406.00      0.00\n",
       "25%    41241.50  54230.00\n",
       "50%    75568.50  84711.00\n",
       "75%   128483.00 139333.00\n",
       "max   170348.00 172792.00"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([fraud.Time.describe(), normal.Time.describe()], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the time feature\n",
    "plt.figure(figsize=(10,8))\n",
    "\n",
    "plt.subplot(2, 2, 1)\n",
    "plt.title('Time Distribution (Seconds)')\n",
    "\n",
    "sns.distplot(data['Time'], color='blue');\n",
    "\n",
    "#plot the amount feature\n",
    "plt.subplot(2, 2, 2)\n",
    "plt.title('Distribution of Amount')\n",
    "sns.distplot(data['Amount'],color='blue');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fee3761a860>"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# data[data.Class == 0].Time.hist(bins=35, color='blue', alpha=0.6)\n",
    "plt.figure(figsize=(12, 10))\n",
    "\n",
    "plt.subplot(2, 2, 1)\n",
    "data[data.Class == 1].Time.hist(bins=35, color='blue', alpha=0.6, label=\"Fraudulant Transaction\")\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(2, 2, 2)\n",
    "data[data.Class == 0].Time.hist(bins=35, color='blue', alpha=0.6, label=\"Non Fraudulant Transaction\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 36 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.hist(figsize=(20, 20));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# heatmap to find any high correlations\n",
    "\n",
    "plt.figure(figsize=(10,10))\n",
    "sns.heatmap(data=data.corr(), cmap=\"seismic\")\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Highest correlations come from:\n",
    "    - Time & V3 (-0.42)\n",
    "    - Amount & V2 (-0.53)\n",
    "    - Amount & V4 (0.4)\n",
    "\n",
    "- While these correlations are high, I don't expect it to run the risk of multicollinearity.\n",
    "\n",
    "- The correlation matrix shows also that none of the V1 to V28 PCA components have any correlation to each other however if we observe Class has some form positive and negative correlations with the V components but has no correlation with Time and Amount."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Data Pre-processing\n",
    "\n",
    "`Time` and `Amount` should be scaled as the other columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fraudulant transaction weight: 0.0017994745785028623\n",
      "Non-Fraudulant transaction weight: 0.9982005254214972\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "scalar = StandardScaler()\n",
    "\n",
    "X = data.drop('Class', axis=1)\n",
    "y = data.Class\n",
    "\n",
    "X_train_v, X_test, y_train_v, y_test = train_test_split(X, y, \n",
    "                                                    test_size=0.3, random_state=42)\n",
    "X_train, X_validate, y_train, y_validate = train_test_split(X_train_v, y_train_v, \n",
    "                                                            test_size=0.2, random_state=42)\n",
    "\n",
    "X_train = scalar.fit_transform(X_train)\n",
    "X_validate = scalar.transform(X_validate)\n",
    "X_test = scalar.transform(X_test)\n",
    "\n",
    "w_p = y_train.value_counts()[0] / len(y_train)\n",
    "w_n = y_train.value_counts()[1] / len(y_train)\n",
    "\n",
    "print(f\"Fraudulant transaction weight: {w_n}\")\n",
    "print(f\"Non-Fraudulant transaction weight: {w_p}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TRAINING: X_train: (159491, 30), y_train: (159491,)\n",
      "_______________________________________________________\n",
      "VALIDATION: X_validate: (39873, 30), y_validate: (39873,)\n",
      "__________________________________________________\n",
      "TESTING: X_test: (85443, 30), y_test: (85443,)\n"
     ]
    }
   ],
   "source": [
    "print(f\"TRAINING: X_train: {X_train.shape}, y_train: {y_train.shape}\\n{'_'*55}\")\n",
    "print(f\"VALIDATION: X_validate: {X_validate.shape}, y_validate: {y_validate.shape}\\n{'_'*50}\")\n",
    "print(f\"TESTING: X_test: {X_test.shape}, y_test: {y_test.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import accuracy_score, confusion_matrix, classification_report, f1_score\n",
    "\n",
    "def print_score(label, prediction, train=True):\n",
    "    if train:\n",
    "        clf_report = pd.DataFrame(classification_report(label, prediction, output_dict=True))\n",
    "        print(\"Train Result:\\n================================================\")\n",
    "        print(f\"Accuracy Score: {accuracy_score(label, prediction) * 100:.2f}%\")\n",
    "        print(\"_______________________________________________\")\n",
    "        print(f\"Classification Report:\\n{clf_report}\")\n",
    "        print(\"_______________________________________________\")\n",
    "        print(f\"Confusion Matrix: \\n {confusion_matrix(y_train, prediction)}\\n\")\n",
    "        \n",
    "    elif train==False:\n",
    "        clf_report = pd.DataFrame(classification_report(label, prediction, output_dict=True))\n",
    "        print(\"Test Result:\\n================================================\")        \n",
    "        print(f\"Accuracy Score: {accuracy_score(label, prediction) * 100:.2f}%\")\n",
    "        print(\"_______________________________________________\")\n",
    "        print(f\"Classification Report:\\n{clf_report}\")\n",
    "        print(\"_______________________________________________\")\n",
    "        print(f\"Confusion Matrix: \\n {confusion_matrix(label, prediction)}\\n\") "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Model Building\n",
    "# 4. 1. Artificial Neural Network (ANNs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_8 (Dense)              (None, 256)               7936      \n",
      "_________________________________________________________________\n",
      "batch_normalization_6 (Batch (None, 256)               1024      \n",
      "_________________________________________________________________\n",
      "dropout_6 (Dropout)          (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_9 (Dense)              (None, 256)               65792     \n",
      "_________________________________________________________________\n",
      "batch_normalization_7 (Batch (None, 256)               1024      \n",
      "_________________________________________________________________\n",
      "dropout_7 (Dropout)          (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_10 (Dense)             (None, 256)               65792     \n",
      "_________________________________________________________________\n",
      "batch_normalization_8 (Batch (None, 256)               1024      \n",
      "_________________________________________________________________\n",
      "dropout_8 (Dropout)          (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_11 (Dense)             (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 142,849\n",
      "Trainable params: 141,313\n",
      "Non-trainable params: 1,536\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from tensorflow import keras\n",
    "\n",
    "model = keras.Sequential([\n",
    "    keras.layers.Dense(256, activation='relu', input_shape=(X_train.shape[-1],)),\n",
    "    keras.layers.BatchNormalization(),\n",
    "    keras.layers.Dropout(0.3),\n",
    "    keras.layers.Dense(256, activation='relu'),\n",
    "    keras.layers.BatchNormalization(),\n",
    "    keras.layers.Dropout(0.3),\n",
    "    keras.layers.Dense(256, activation='relu'),\n",
    "    keras.layers.BatchNormalization(),\n",
    "    keras.layers.Dropout(0.3),\n",
    "    keras.layers.Dense(1, activation='sigmoid'),\n",
    "])\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "_kg_hide-output": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 159491 samples, validate on 39873 samples\n",
      "Epoch 1/300\n",
      "159491/159491 [==============================] - 3s 21us/sample - loss: 0.4859 - fn: 56.0000 - fp: 28702.0000 - tn: 130502.0000 - tp: 231.0000 - precision: 0.0080 - recall: 0.8049 - val_loss: 0.1705 - val_fn: 14.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 55.0000 - val_precision: 0.7971 - val_recall: 0.7971\n",
      "Epoch 2/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 0.1043 - fn: 74.0000 - fp: 544.0000 - tn: 158660.0000 - tp: 213.0000 - precision: 0.2814 - recall: 0.7422 - val_loss: 0.0355 - val_fn: 14.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 55.0000 - val_precision: 0.7971 - val_recall: 0.7971\n",
      "Epoch 3/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0363 - fn: 103.0000 - fp: 124.0000 - tn: 159080.0000 - tp: 184.0000 - precision: 0.5974 - recall: 0.6411 - val_loss: 0.0212 - val_fn: 14.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 55.0000 - val_precision: 0.7971 - val_recall: 0.7971\n",
      "Epoch 4/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0203 - fn: 90.0000 - fp: 99.0000 - tn: 159105.0000 - tp: 197.0000 - precision: 0.6655 - recall: 0.6864 - val_loss: 0.0169 - val_fn: 14.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 55.0000 - val_precision: 0.7971 - val_recall: 0.7971\n",
      "Epoch 5/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0195 - fn: 108.0000 - fp: 74.0000 - tn: 159130.0000 - tp: 179.0000 - precision: 0.7075 - recall: 0.6237 - val_loss: 0.0173 - val_fn: 13.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 56.0000 - val_precision: 0.8000 - val_recall: 0.8116\n",
      "Epoch 6/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0163 - fn: 93.0000 - fp: 68.0000 - tn: 159136.0000 - tp: 194.0000 - precision: 0.7405 - recall: 0.6760 - val_loss: 0.0192 - val_fn: 13.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 56.0000 - val_precision: 0.8000 - val_recall: 0.8116\n",
      "Epoch 7/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0141 - fn: 93.0000 - fp: 53.0000 - tn: 159151.0000 - tp: 194.0000 - precision: 0.7854 - recall: 0.6760 - val_loss: 0.0184 - val_fn: 13.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 56.0000 - val_precision: 0.8000 - val_recall: 0.8116\n",
      "Epoch 8/300\n",
      "159491/159491 [==============================] - 1s 7us/sample - loss: 0.0154 - fn: 90.0000 - fp: 47.0000 - tn: 159157.0000 - tp: 197.0000 - precision: 0.8074 - recall: 0.6864 - val_loss: 0.0116 - val_fn: 13.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 56.0000 - val_precision: 0.8000 - val_recall: 0.8116\n",
      "Epoch 9/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0115 - fn: 93.0000 - fp: 40.0000 - tn: 159164.0000 - tp: 194.0000 - precision: 0.8291 - recall: 0.6760 - val_loss: 0.0186 - val_fn: 13.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 56.0000 - val_precision: 0.8000 - val_recall: 0.8116\n",
      "Epoch 10/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0147 - fn: 93.0000 - fp: 50.0000 - tn: 159154.0000 - tp: 194.0000 - precision: 0.7951 - recall: 0.6760 - val_loss: 0.0183 - val_fn: 13.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 56.0000 - val_precision: 0.8000 - val_recall: 0.8116\n",
      "Epoch 11/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0104 - fn: 87.0000 - fp: 42.0000 - tn: 159162.0000 - tp: 200.0000 - precision: 0.8264 - recall: 0.6969 - val_loss: 0.0114 - val_fn: 13.0000 - val_fp: 14.0000 - val_tn: 39790.0000 - val_tp: 56.0000 - val_precision: 0.8000 - val_recall: 0.8116\n",
      "Epoch 12/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0084 - fn: 89.0000 - fp: 40.0000 - tn: 159164.0000 - tp: 198.0000 - precision: 0.8319 - recall: 0.6899 - val_loss: 0.0063 - val_fn: 16.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 53.0000 - val_precision: 0.8154 - val_recall: 0.7681\n",
      "Epoch 13/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0056 - fn: 86.0000 - fp: 25.0000 - tn: 159179.0000 - tp: 201.0000 - precision: 0.8894 - recall: 0.7003 - val_loss: 0.0052 - val_fn: 23.0000 - val_fp: 3.0000 - val_tn: 39801.0000 - val_tp: 46.0000 - val_precision: 0.9388 - val_recall: 0.6667\n",
      "Epoch 14/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0059 - fn: 96.0000 - fp: 37.0000 - tn: 159167.0000 - tp: 191.0000 - precision: 0.8377 - recall: 0.6655 - val_loss: 0.0061 - val_fn: 13.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 56.0000 - val_precision: 0.8235 - val_recall: 0.8116\n",
      "Epoch 15/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0049 - fn: 86.0000 - fp: 27.0000 - tn: 159177.0000 - tp: 201.0000 - precision: 0.8816 - recall: 0.7003 - val_loss: 0.0059 - val_fn: 13.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 56.0000 - val_precision: 0.8235 - val_recall: 0.8116\n",
      "Epoch 16/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0048 - fn: 83.0000 - fp: 31.0000 - tn: 159173.0000 - tp: 204.0000 - precision: 0.8681 - recall: 0.7108 - val_loss: 0.0051 - val_fn: 16.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 53.0000 - val_precision: 0.8548 - val_recall: 0.7681\n",
      "Epoch 17/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0045 - fn: 84.0000 - fp: 25.0000 - tn: 159179.0000 - tp: 203.0000 - precision: 0.8904 - recall: 0.7073 - val_loss: 0.0046 - val_fn: 18.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 51.0000 - val_precision: 0.9107 - val_recall: 0.7391\n",
      "Epoch 18/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0038 - fn: 81.0000 - fp: 27.0000 - tn: 159177.0000 - tp: 206.0000 - precision: 0.8841 - recall: 0.7178 - val_loss: 0.0048 - val_fn: 17.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 52.0000 - val_precision: 0.8966 - val_recall: 0.7536\n",
      "Epoch 19/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0036 - fn: 80.0000 - fp: 26.0000 - tn: 159178.0000 - tp: 207.0000 - precision: 0.8884 - recall: 0.7213 - val_loss: 0.0048 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 20/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0036 - fn: 79.0000 - fp: 26.0000 - tn: 159178.0000 - tp: 208.0000 - precision: 0.8889 - recall: 0.7247 - val_loss: 0.0045 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 21/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0032 - fn: 79.0000 - fp: 23.0000 - tn: 159181.0000 - tp: 208.0000 - precision: 0.9004 - recall: 0.7247 - val_loss: 0.0045 - val_fn: 17.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 52.0000 - val_precision: 0.8814 - val_recall: 0.7536\n",
      "Epoch 22/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0030 - fn: 70.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 217.0000 - precision: 0.9274 - recall: 0.7561 - val_loss: 0.0045 - val_fn: 17.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 52.0000 - val_precision: 0.8667 - val_recall: 0.7536\n",
      "Epoch 23/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 0.0032 - fn: 80.0000 - fp: 21.0000 - tn: 159183.0000 - tp: 207.0000 - precision: 0.9079 - recall: 0.7213 - val_loss: 0.0046 - val_fn: 17.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 52.0000 - val_precision: 0.8966 - val_recall: 0.7536\n",
      "Epoch 24/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 0.0031 - fn: 72.0000 - fp: 20.0000 - tn: 159184.0000 - tp: 215.0000 - precision: 0.9149 - recall: 0.7491 - val_loss: 0.0053 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 25/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0030 - fn: 73.0000 - fp: 24.0000 - tn: 159180.0000 - tp: 214.0000 - precision: 0.8992 - recall: 0.7456 - val_loss: 0.0047 - val_fn: 19.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 50.0000 - val_precision: 0.9091 - val_recall: 0.7246\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 26/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0030 - fn: 64.0000 - fp: 24.0000 - tn: 159180.0000 - tp: 223.0000 - precision: 0.9028 - recall: 0.7770 - val_loss: 0.0045 - val_fn: 17.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 52.0000 - val_precision: 0.8966 - val_recall: 0.7536\n",
      "Epoch 27/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0028 - fn: 68.0000 - fp: 15.0000 - tn: 159189.0000 - tp: 219.0000 - precision: 0.9359 - recall: 0.7631 - val_loss: 0.0046 - val_fn: 20.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 49.0000 - val_precision: 0.9074 - val_recall: 0.7101\n",
      "Epoch 28/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0027 - fn: 74.0000 - fp: 21.0000 - tn: 159183.0000 - tp: 213.0000 - precision: 0.9103 - recall: 0.7422 - val_loss: 0.0049 - val_fn: 29.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 40.0000 - val_precision: 0.8889 - val_recall: 0.5797\n",
      "Epoch 29/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0028 - fn: 74.0000 - fp: 19.0000 - tn: 159185.0000 - tp: 213.0000 - precision: 0.9181 - recall: 0.7422 - val_loss: 0.0048 - val_fn: 15.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 54.0000 - val_precision: 0.8438 - val_recall: 0.7826\n",
      "Epoch 30/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0024 - fn: 60.0000 - fp: 18.0000 - tn: 159186.0000 - tp: 227.0000 - precision: 0.9265 - recall: 0.7909 - val_loss: 0.0046 - val_fn: 18.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 51.0000 - val_precision: 0.8793 - val_recall: 0.7391\n",
      "Epoch 31/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0026 - fn: 68.0000 - fp: 22.0000 - tn: 159182.0000 - tp: 219.0000 - precision: 0.9087 - recall: 0.7631 - val_loss: 0.0045 - val_fn: 17.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 52.0000 - val_precision: 0.9123 - val_recall: 0.7536\n",
      "Epoch 32/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0025 - fn: 65.0000 - fp: 16.0000 - tn: 159188.0000 - tp: 222.0000 - precision: 0.9328 - recall: 0.7735 - val_loss: 0.0048 - val_fn: 13.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 56.0000 - val_precision: 0.8235 - val_recall: 0.8116\n",
      "Epoch 33/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0024 - fn: 64.0000 - fp: 18.0000 - tn: 159186.0000 - tp: 223.0000 - precision: 0.9253 - recall: 0.7770 - val_loss: 0.0050 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 34/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0022 - fn: 62.0000 - fp: 16.0000 - tn: 159188.0000 - tp: 225.0000 - precision: 0.9336 - recall: 0.7840 - val_loss: 0.0048 - val_fn: 17.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 52.0000 - val_precision: 0.8667 - val_recall: 0.7536\n",
      "Epoch 35/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0024 - fn: 61.0000 - fp: 21.0000 - tn: 159183.0000 - tp: 226.0000 - precision: 0.9150 - recall: 0.7875 - val_loss: 0.0045 - val_fn: 19.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 50.0000 - val_precision: 0.8621 - val_recall: 0.7246\n",
      "Epoch 36/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0024 - fn: 58.0000 - fp: 20.0000 - tn: 159184.0000 - tp: 229.0000 - precision: 0.9197 - recall: 0.7979 - val_loss: 0.0047 - val_fn: 19.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 50.0000 - val_precision: 0.9091 - val_recall: 0.7246\n",
      "Epoch 37/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0025 - fn: 60.0000 - fp: 18.0000 - tn: 159186.0000 - tp: 227.0000 - precision: 0.9265 - recall: 0.7909 - val_loss: 0.0049 - val_fn: 17.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 52.0000 - val_precision: 0.8966 - val_recall: 0.7536\n",
      "Epoch 38/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0022 - fn: 62.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 225.0000 - precision: 0.9298 - recall: 0.7840 - val_loss: 0.0049 - val_fn: 16.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 53.0000 - val_precision: 0.8833 - val_recall: 0.7681\n",
      "Epoch 39/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 0.0023 - fn: 55.0000 - fp: 20.0000 - tn: 159184.0000 - tp: 232.0000 - precision: 0.9206 - recall: 0.8084 - val_loss: 0.0049 - val_fn: 16.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 53.0000 - val_precision: 0.8833 - val_recall: 0.7681\n",
      "Epoch 40/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 0.0021 - fn: 62.0000 - fp: 20.0000 - tn: 159184.0000 - tp: 225.0000 - precision: 0.9184 - recall: 0.7840 - val_loss: 0.0050 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 41/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0022 - fn: 56.0000 - fp: 23.0000 - tn: 159181.0000 - tp: 231.0000 - precision: 0.9094 - recall: 0.8049 - val_loss: 0.0048 - val_fn: 17.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 52.0000 - val_precision: 0.8667 - val_recall: 0.7536\n",
      "Epoch 42/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0020 - fn: 57.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 230.0000 - precision: 0.9312 - recall: 0.8014 - val_loss: 0.0045 - val_fn: 15.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 54.0000 - val_precision: 0.9000 - val_recall: 0.7826\n",
      "Epoch 43/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0018 - fn: 48.0000 - fp: 16.0000 - tn: 159188.0000 - tp: 239.0000 - precision: 0.9373 - recall: 0.8328 - val_loss: 0.0050 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 44/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0017 - fn: 55.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 232.0000 - precision: 0.9431 - recall: 0.8084 - val_loss: 0.0048 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 45/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0020 - fn: 52.0000 - fp: 19.0000 - tn: 159185.0000 - tp: 235.0000 - precision: 0.9252 - recall: 0.8188 - val_loss: 0.0053 - val_fn: 13.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 56.0000 - val_precision: 0.8889 - val_recall: 0.8116\n",
      "Epoch 46/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0015 - fn: 45.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 242.0000 - precision: 0.9565 - recall: 0.8432 - val_loss: 0.0055 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 47/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0019 - fn: 53.0000 - fp: 22.0000 - tn: 159182.0000 - tp: 234.0000 - precision: 0.9141 - recall: 0.8153 - val_loss: 0.0046 - val_fn: 18.0000 - val_fp: 4.0000 - val_tn: 39800.0000 - val_tp: 51.0000 - val_precision: 0.9273 - val_recall: 0.7391\n",
      "Epoch 48/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0020 - fn: 53.0000 - fp: 16.0000 - tn: 159188.0000 - tp: 234.0000 - precision: 0.9360 - recall: 0.8153 - val_loss: 0.0053 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 49/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0017 - fn: 54.0000 - fp: 15.0000 - tn: 159189.0000 - tp: 233.0000 - precision: 0.9395 - recall: 0.8118 - val_loss: 0.0050 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 50/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0017 - fn: 51.0000 - fp: 18.0000 - tn: 159186.0000 - tp: 236.0000 - precision: 0.9291 - recall: 0.8223 - val_loss: 0.0050 - val_fn: 24.0000 - val_fp: 4.0000 - val_tn: 39800.0000 - val_tp: 45.0000 - val_precision: 0.9184 - val_recall: 0.6522\n",
      "Epoch 51/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0017 - fn: 57.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 230.0000 - precision: 0.9312 - recall: 0.8014 - val_loss: 0.0050 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 52/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0016 - fn: 53.0000 - fp: 16.0000 - tn: 159188.0000 - tp: 234.0000 - precision: 0.9360 - recall: 0.8153 - val_loss: 0.0046 - val_fn: 16.0000 - val_fp: 4.0000 - val_tn: 39800.0000 - val_tp: 53.0000 - val_precision: 0.9298 - val_recall: 0.7681\n",
      "Epoch 53/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0017 - fn: 54.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 233.0000 - precision: 0.9433 - recall: 0.8118 - val_loss: 0.0051 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 54/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0015 - fn: 52.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 235.0000 - precision: 0.9553 - recall: 0.8188 - val_loss: 0.0054 - val_fn: 13.0000 - val_fp: 11.0000 - val_tn: 39793.0000 - val_tp: 56.0000 - val_precision: 0.8358 - val_recall: 0.8116\n",
      "Epoch 55/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 0.0016 - fn: 44.0000 - fp: 16.0000 - tn: 159188.0000 - tp: 243.0000 - precision: 0.9382 - recall: 0.8467 - val_loss: 0.0050 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 56/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0015 - fn: 48.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 239.0000 - precision: 0.9336 - recall: 0.8328 - val_loss: 0.0052 - val_fn: 13.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 56.0000 - val_precision: 0.8235 - val_recall: 0.8116\n",
      "Epoch 57/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0017 - fn: 50.0000 - fp: 21.0000 - tn: 159183.0000 - tp: 237.0000 - precision: 0.9186 - recall: 0.8258 - val_loss: 0.0054 - val_fn: 13.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 56.0000 - val_precision: 0.8235 - val_recall: 0.8116\n",
      "Epoch 58/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0016 - fn: 40.0000 - fp: 15.0000 - tn: 159189.0000 - tp: 247.0000 - precision: 0.9427 - recall: 0.8606 - val_loss: 0.0052 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 59/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0012 - fn: 44.0000 - fp: 20.0000 - tn: 159184.0000 - tp: 243.0000 - precision: 0.9240 - recall: 0.8467 - val_loss: 0.0054 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 60/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0015 - fn: 46.0000 - fp: 22.0000 - tn: 159182.0000 - tp: 241.0000 - precision: 0.9163 - recall: 0.8397 - val_loss: 0.0050 - val_fn: 13.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 56.0000 - val_precision: 0.8889 - val_recall: 0.8116\n",
      "Epoch 61/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0013 - fn: 45.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 242.0000 - precision: 0.9490 - recall: 0.8432 - val_loss: 0.0051 - val_fn: 15.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 54.0000 - val_precision: 0.9000 - val_recall: 0.7826\n",
      "Epoch 62/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0014 - fn: 42.0000 - fp: 15.0000 - tn: 159189.0000 - tp: 245.0000 - precision: 0.9423 - recall: 0.8537 - val_loss: 0.0050 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 63/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0012 - fn: 40.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 247.0000 - precision: 0.9537 - recall: 0.8606 - val_loss: 0.0048 - val_fn: 17.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 52.0000 - val_precision: 0.8814 - val_recall: 0.7536\n",
      "Epoch 64/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0014 - fn: 47.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 240.0000 - precision: 0.9524 - recall: 0.8362 - val_loss: 0.0052 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 65/300\n",
      "159491/159491 [==============================] - 1s 8us/sample - loss: 0.0013 - fn: 42.0000 - fp: 15.0000 - tn: 159189.0000 - tp: 245.0000 - precision: 0.9423 - recall: 0.8537 - val_loss: 0.0052 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 66/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0012 - fn: 39.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 248.0000 - precision: 0.9502 - recall: 0.8641 - val_loss: 0.0057 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 67/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0014 - fn: 41.0000 - fp: 15.0000 - tn: 159189.0000 - tp: 246.0000 - precision: 0.9425 - recall: 0.8571 - val_loss: 0.0066 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 68/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 0.0015 - fn: 43.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 244.0000 - precision: 0.9531 - recall: 0.8502 - val_loss: 0.0066 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 69/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 0.0014 - fn: 40.0000 - fp: 18.0000 - tn: 159186.0000 - tp: 247.0000 - precision: 0.9321 - recall: 0.8606 - val_loss: 0.0060 - val_fn: 18.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 51.0000 - val_precision: 0.9107 - val_recall: 0.7391\n",
      "Epoch 70/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 0.0013 - fn: 43.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 244.0000 - precision: 0.9457 - recall: 0.8502 - val_loss: 0.0059 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 71/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0011 - fn: 39.0000 - fp: 16.0000 - tn: 159188.0000 - tp: 248.0000 - precision: 0.9394 - recall: 0.8641 - val_loss: 0.0059 - val_fn: 16.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 53.0000 - val_precision: 0.8689 - val_recall: 0.7681\n",
      "Epoch 72/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0013 - fn: 32.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 255.0000 - precision: 0.9480 - recall: 0.8885 - val_loss: 0.0057 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 73/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0011 - fn: 33.0000 - fp: 19.0000 - tn: 159185.0000 - tp: 254.0000 - precision: 0.9304 - recall: 0.8850 - val_loss: 0.0055 - val_fn: 15.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 54.0000 - val_precision: 0.9000 - val_recall: 0.7826\n",
      "Epoch 74/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0011 - fn: 38.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 249.0000 - precision: 0.9651 - recall: 0.8676 - val_loss: 0.0059 - val_fn: 13.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 56.0000 - val_precision: 0.8889 - val_recall: 0.8116\n",
      "Epoch 75/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0014 - fn: 41.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 246.0000 - precision: 0.9354 - recall: 0.8571 - val_loss: 0.0064 - val_fn: 20.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 49.0000 - val_precision: 0.9074 - val_recall: 0.7101\n",
      "Epoch 76/300\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0013 - fn: 37.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 250.0000 - precision: 0.9363 - recall: 0.8711 - val_loss: 0.0056 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 77/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0010 - fn: 31.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 256.0000 - precision: 0.9517 - recall: 0.8920 - val_loss: 0.0059 - val_fn: 17.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 52.0000 - val_precision: 0.8814 - val_recall: 0.7536\n",
      "Epoch 78/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0011 - fn: 33.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 254.0000 - precision: 0.9478 - recall: 0.8850 - val_loss: 0.0058 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 79/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0010 - fn: 28.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 259.0000 - precision: 0.9522 - recall: 0.9024 - val_loss: 0.0060 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 80/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0011 - fn: 38.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 249.0000 - precision: 0.9468 - recall: 0.8676 - val_loss: 0.0058 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 81/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 9.6854e-04 - fn: 33.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 254.0000 - precision: 0.9478 - recall: 0.8850 - val_loss: 0.0054 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 82/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 8.8699e-04 - fn: 33.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 254.0000 - precision: 0.9513 - recall: 0.8850 - val_loss: 0.0065 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 83/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0010 - fn: 33.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 254.0000 - precision: 0.9478 - recall: 0.8850 - val_loss: 0.0065 - val_fn: 16.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 53.0000 - val_precision: 0.8548 - val_recall: 0.7681\n",
      "Epoch 84/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 9.8176e-04 - fn: 28.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 259.0000 - precision: 0.9522 - recall: 0.9024 - val_loss: 0.0058 - val_fn: 15.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 54.0000 - val_precision: 0.9153 - val_recall: 0.7826\n",
      "Epoch 85/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 0.0012 - fn: 38.0000 - fp: 20.0000 - tn: 159184.0000 - tp: 249.0000 - precision: 0.9257 - recall: 0.8676 - val_loss: 0.0061 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 86/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 0.0010 - fn: 36.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 251.0000 - precision: 0.9366 - recall: 0.8746 - val_loss: 0.0062 - val_fn: 16.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 53.0000 - val_precision: 0.8833 - val_recall: 0.7681\n",
      "Epoch 87/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 9.6321e-04 - fn: 33.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 254.0000 - precision: 0.9513 - recall: 0.8850 - val_loss: 0.0062 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 88/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 8.5205e-04 - fn: 30.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 257.0000 - precision: 0.9483 - recall: 0.8955 - val_loss: 0.0067 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 89/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0012 - fn: 36.0000 - fp: 22.0000 - tn: 159182.0000 - tp: 251.0000 - precision: 0.9194 - recall: 0.8746 - val_loss: 0.0070 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 90/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 9.0820e-04 - fn: 29.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 258.0000 - precision: 0.9591 - recall: 0.8990 - val_loss: 0.0067 - val_fn: 16.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 53.0000 - val_precision: 0.8833 - val_recall: 0.7681\n",
      "Epoch 91/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0010 - fn: 36.0000 - fp: 17.0000 - tn: 159187.0000 - tp: 251.0000 - precision: 0.9366 - recall: 0.8746 - val_loss: 0.0064 - val_fn: 17.0000 - val_fp: 4.0000 - val_tn: 39800.0000 - val_tp: 52.0000 - val_precision: 0.9286 - val_recall: 0.7536\n",
      "Epoch 92/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 0.0010 - fn: 34.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 253.0000 - precision: 0.9547 - recall: 0.8815 - val_loss: 0.0064 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 93/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 9.1671e-04 - fn: 28.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 259.0000 - precision: 0.9522 - recall: 0.9024 - val_loss: 0.0063 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 94/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.7722e-04 - fn: 22.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 265.0000 - precision: 0.9532 - recall: 0.9233 - val_loss: 0.0064 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 95/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.7799e-04 - fn: 24.0000 - fp: 14.0000 - tn: 159190.0000 - tp: 263.0000 - precision: 0.9495 - recall: 0.9164 - val_loss: 0.0061 - val_fn: 13.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 56.0000 - val_precision: 0.8889 - val_recall: 0.8116\n",
      "Epoch 96/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 8.9702e-04 - fn: 32.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 255.0000 - precision: 0.9515 - recall: 0.8885 - val_loss: 0.0059 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 97/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.4815e-04 - fn: 24.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 263.0000 - precision: 0.9669 - recall: 0.9164 - val_loss: 0.0062 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 98/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.7793e-04 - fn: 25.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 262.0000 - precision: 0.9740 - recall: 0.9129 - val_loss: 0.0062 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 99/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 8.1968e-04 - fn: 22.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 265.0000 - precision: 0.9601 - recall: 0.9233 - val_loss: 0.0064 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 100/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.8394e-04 - fn: 24.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 263.0000 - precision: 0.9529 - recall: 0.9164 - val_loss: 0.0060 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 101/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 7.7056e-04 - fn: 24.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 263.0000 - precision: 0.9529 - recall: 0.9164 - val_loss: 0.0057 - val_fn: 16.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 53.0000 - val_precision: 0.9138 - val_recall: 0.7681\n",
      "Epoch 102/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 8.5635e-04 - fn: 28.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 259.0000 - precision: 0.9522 - recall: 0.9024 - val_loss: 0.0058 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 103/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.9294e-04 - fn: 29.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 258.0000 - precision: 0.9591 - recall: 0.8990 - val_loss: 0.0061 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 104/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 8.3591e-04 - fn: 24.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 263.0000 - precision: 0.9529 - recall: 0.9164 - val_loss: 0.0059 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 105/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.7620e-04 - fn: 23.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 264.0000 - precision: 0.9531 - recall: 0.9199 - val_loss: 0.0062 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 106/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.3468e-04 - fn: 27.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 260.0000 - precision: 0.9701 - recall: 0.9059 - val_loss: 0.0063 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 107/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.2148e-04 - fn: 25.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 262.0000 - precision: 0.9527 - recall: 0.9129 - val_loss: 0.0063 - val_fn: 16.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 53.0000 - val_precision: 0.8983 - val_recall: 0.7681\n",
      "Epoch 108/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 8.7943e-04 - fn: 26.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 261.0000 - precision: 0.9560 - recall: 0.9094 - val_loss: 0.0062 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 109/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.2424e-04 - fn: 27.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 260.0000 - precision: 0.9738 - recall: 0.9059 - val_loss: 0.0064 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 110/300\n",
      "159491/159491 [==============================] - 1s 8us/sample - loss: 0.0011 - fn: 30.0000 - fp: 18.0000 - tn: 159186.0000 - tp: 257.0000 - precision: 0.9345 - recall: 0.8955 - val_loss: 0.0064 - val_fn: 15.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 54.0000 - val_precision: 0.9000 - val_recall: 0.7826\n",
      "Epoch 111/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.9243e-04 - fn: 29.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 258.0000 - precision: 0.9627 - recall: 0.8990 - val_loss: 0.0065 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 112/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.3825e-04 - fn: 19.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 268.0000 - precision: 0.9675 - recall: 0.9338 - val_loss: 0.0063 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 113/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.5662e-04 - fn: 19.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 268.0000 - precision: 0.9640 - recall: 0.9338 - val_loss: 0.0064 - val_fn: 13.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 56.0000 - val_precision: 0.8889 - val_recall: 0.8116\n",
      "Epoch 114/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.6240e-04 - fn: 25.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 262.0000 - precision: 0.9527 - recall: 0.9129 - val_loss: 0.0064 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 115/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.8147e-04 - fn: 20.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 267.0000 - precision: 0.9639 - recall: 0.9303 - val_loss: 0.0068 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 116/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 7.4114e-04 - fn: 23.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 264.0000 - precision: 0.9600 - recall: 0.9199 - val_loss: 0.0062 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 117/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 6.5416e-04 - fn: 15.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 272.0000 - precision: 0.9680 - recall: 0.9477 - val_loss: 0.0062 - val_fn: 15.0000 - val_fp: 11.0000 - val_tn: 39793.0000 - val_tp: 54.0000 - val_precision: 0.8308 - val_recall: 0.7826\n",
      "Epoch 118/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.9064e-04 - fn: 22.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 265.0000 - precision: 0.9601 - recall: 0.9233 - val_loss: 0.0064 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 119/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.6247e-04 - fn: 20.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 267.0000 - precision: 0.9674 - recall: 0.9303 - val_loss: 0.0064 - val_fn: 13.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 56.0000 - val_precision: 0.8235 - val_recall: 0.8116\n",
      "Epoch 120/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.7345e-04 - fn: 18.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 269.0000 - precision: 0.9607 - recall: 0.9373 - val_loss: 0.0064 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 121/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.7623e-04 - fn: 19.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 268.0000 - precision: 0.9640 - recall: 0.9338 - val_loss: 0.0062 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 122/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.2076e-04 - fn: 19.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 268.0000 - precision: 0.9710 - recall: 0.9338 - val_loss: 0.0063 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 123/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.9264e-04 - fn: 21.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 266.0000 - precision: 0.9638 - recall: 0.9268 - val_loss: 0.0073 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 124/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 8.6501e-04 - fn: 21.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 266.0000 - precision: 0.9568 - recall: 0.9268 - val_loss: 0.0063 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 125/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.9235e-04 - fn: 23.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 264.0000 - precision: 0.9635 - recall: 0.9199 - val_loss: 0.0065 - val_fn: 20.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 49.0000 - val_precision: 0.8750 - val_recall: 0.7101\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 126/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.2551e-04 - fn: 22.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 265.0000 - precision: 0.9743 - recall: 0.9233 - val_loss: 0.0066 - val_fn: 14.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 55.0000 - val_precision: 0.9167 - val_recall: 0.7971\n",
      "Epoch 127/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 5.7184e-04 - fn: 21.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 266.0000 - precision: 0.9638 - recall: 0.9268 - val_loss: 0.0066 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 128/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.6659e-04 - fn: 16.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 271.0000 - precision: 0.9713 - recall: 0.9443 - val_loss: 0.0070 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 129/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.5194e-04 - fn: 17.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 270.0000 - precision: 0.9783 - recall: 0.9408 - val_loss: 0.0067 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 130/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.4185e-04 - fn: 18.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 269.0000 - precision: 0.9711 - recall: 0.9373 - val_loss: 0.0070 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 131/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.5417e-04 - fn: 17.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 270.0000 - precision: 0.9609 - recall: 0.9408 - val_loss: 0.0067 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 132/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 5.2499e-04 - fn: 18.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 269.0000 - precision: 0.9676 - recall: 0.9373 - val_loss: 0.0065 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 133/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.6890e-04 - fn: 14.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 273.0000 - precision: 0.9613 - recall: 0.9512 - val_loss: 0.0060 - val_fn: 15.0000 - val_fp: 4.0000 - val_tn: 39800.0000 - val_tp: 54.0000 - val_precision: 0.9310 - val_recall: 0.7826\n",
      "Epoch 134/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.7169e-04 - fn: 23.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 264.0000 - precision: 0.9531 - recall: 0.9199 - val_loss: 0.0066 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 135/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.3518e-04 - fn: 18.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 269.0000 - precision: 0.9676 - recall: 0.9373 - val_loss: 0.0066 - val_fn: 21.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 48.0000 - val_precision: 0.8727 - val_recall: 0.6957\n",
      "Epoch 136/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.7904e-04 - fn: 20.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 267.0000 - precision: 0.9604 - recall: 0.9303 - val_loss: 0.0066 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 137/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.8278e-04 - fn: 19.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 268.0000 - precision: 0.9710 - recall: 0.9338 - val_loss: 0.0067 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 138/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.2436e-04 - fn: 20.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 267.0000 - precision: 0.9639 - recall: 0.9303 - val_loss: 0.0069 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 139/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.7755e-04 - fn: 16.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 271.0000 - precision: 0.9610 - recall: 0.9443 - val_loss: 0.0066 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 140/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.1207e-04 - fn: 19.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 268.0000 - precision: 0.9675 - recall: 0.9338 - val_loss: 0.0065 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 141/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 7.3742e-04 - fn: 20.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 267.0000 - precision: 0.9674 - recall: 0.9303 - val_loss: 0.0063 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 142/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.2673e-04 - fn: 25.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 262.0000 - precision: 0.9527 - recall: 0.9129 - val_loss: 0.0062 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 143/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.0601e-04 - fn: 15.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 272.0000 - precision: 0.9714 - recall: 0.9477 - val_loss: 0.0069 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 144/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.5733e-04 - fn: 18.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 269.0000 - precision: 0.9818 - recall: 0.9373 - val_loss: 0.0064 - val_fn: 14.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 55.0000 - val_precision: 0.9167 - val_recall: 0.7971\n",
      "Epoch 145/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.3996e-04 - fn: 18.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 269.0000 - precision: 0.9642 - recall: 0.9373 - val_loss: 0.0072 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 146/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.6741e-04 - fn: 19.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 268.0000 - precision: 0.9710 - recall: 0.9338 - val_loss: 0.0071 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 147/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 5.8892e-04 - fn: 19.0000 - fp: 13.0000 - tn: 159191.0000 - tp: 268.0000 - precision: 0.9537 - recall: 0.9338 - val_loss: 0.0067 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 148/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 4.1562e-04 - fn: 13.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 274.0000 - precision: 0.9821 - recall: 0.9547 - val_loss: 0.0069 - val_fn: 14.0000 - val_fp: 11.0000 - val_tn: 39793.0000 - val_tp: 55.0000 - val_precision: 0.8333 - val_recall: 0.7971\n",
      "Epoch 149/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.4916e-04 - fn: 15.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 272.0000 - precision: 0.9645 - recall: 0.9477 - val_loss: 0.0066 - val_fn: 16.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 53.0000 - val_precision: 0.8689 - val_recall: 0.7681\n",
      "Epoch 150/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.8643e-04 - fn: 18.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 269.0000 - precision: 0.9818 - recall: 0.9373 - val_loss: 0.0063 - val_fn: 16.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 53.0000 - val_precision: 0.8833 - val_recall: 0.7681\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 151/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.9819e-04 - fn: 20.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 267.0000 - precision: 0.9604 - recall: 0.9303 - val_loss: 0.0066 - val_fn: 14.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 55.0000 - val_precision: 0.8209 - val_recall: 0.7971\n",
      "Epoch 152/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.7220e-04 - fn: 9.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 278.0000 - precision: 0.9823 - recall: 0.9686 - val_loss: 0.0061 - val_fn: 15.0000 - val_fp: 4.0000 - val_tn: 39800.0000 - val_tp: 54.0000 - val_precision: 0.9310 - val_recall: 0.7826\n",
      "Epoch 153/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.2844e-04 - fn: 15.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 272.0000 - precision: 0.9611 - recall: 0.9477 - val_loss: 0.0065 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 154/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.4049e-04 - fn: 17.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 270.0000 - precision: 0.9609 - recall: 0.9408 - val_loss: 0.0062 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 155/300\n",
      "159491/159491 [==============================] - 1s 7us/sample - loss: 4.4462e-04 - fn: 14.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 273.0000 - precision: 0.9820 - recall: 0.9512 - val_loss: 0.0093 - val_fn: 16.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 53.0000 - val_precision: 0.8689 - val_recall: 0.7681\n",
      "Epoch 156/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 8.1759e-04 - fn: 18.0000 - fp: 15.0000 - tn: 159189.0000 - tp: 269.0000 - precision: 0.9472 - recall: 0.9373 - val_loss: 0.0076 - val_fn: 15.0000 - val_fp: 4.0000 - val_tn: 39800.0000 - val_tp: 54.0000 - val_precision: 0.9310 - val_recall: 0.7826\n",
      "Epoch 157/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.6813e-04 - fn: 22.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 265.0000 - precision: 0.9636 - recall: 0.9233 - val_loss: 0.0077 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 158/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.6256e-04 - fn: 15.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 272.0000 - precision: 0.9611 - recall: 0.9477 - val_loss: 0.0076 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 159/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.2024e-04 - fn: 12.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 275.0000 - precision: 0.9752 - recall: 0.9582 - val_loss: 0.0075 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 160/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.3576e-04 - fn: 18.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 269.0000 - precision: 0.9746 - recall: 0.9373 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 161/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.6558e-04 - fn: 11.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 276.0000 - precision: 0.9753 - recall: 0.9617 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 162/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 4.8901e-04 - fn: 14.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 273.0000 - precision: 0.9613 - recall: 0.9512 - val_loss: 0.0079 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 163/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 4.3683e-04 - fn: 13.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 274.0000 - precision: 0.9786 - recall: 0.9547 - val_loss: 0.0078 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 164/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.7351e-04 - fn: 13.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 274.0000 - precision: 0.9856 - recall: 0.9547 - val_loss: 0.0081 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 165/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.7894e-04 - fn: 12.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 275.0000 - precision: 0.9683 - recall: 0.9582 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 166/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.2191e-04 - fn: 15.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 272.0000 - precision: 0.9680 - recall: 0.9477 - val_loss: 0.0078 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 167/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.8416e-04 - fn: 19.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 268.0000 - precision: 0.9571 - recall: 0.9338 - val_loss: 0.0078 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 168/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.2035e-04 - fn: 17.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 270.0000 - precision: 0.9677 - recall: 0.9408 - val_loss: 0.0075 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 169/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.5776e-04 - fn: 14.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 273.0000 - precision: 0.9715 - recall: 0.9512 - val_loss: 0.0075 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 170/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.8820e-04 - fn: 11.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 276.0000 - precision: 0.9718 - recall: 0.9617 - val_loss: 0.0078 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 171/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.6732e-04 - fn: 15.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 272.0000 - precision: 0.9680 - recall: 0.9477 - val_loss: 0.0079 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 172/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.7529e-04 - fn: 16.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 271.0000 - precision: 0.9679 - recall: 0.9443 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 13.0000 - val_tn: 39791.0000 - val_tp: 55.0000 - val_precision: 0.8088 - val_recall: 0.7971\n",
      "Epoch 173/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 6.4442e-04 - fn: 18.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 269.0000 - precision: 0.9642 - recall: 0.9373 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 174/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.6802e-04 - fn: 11.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 276.0000 - precision: 0.9753 - recall: 0.9617 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 175/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.3756e-04 - fn: 15.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 272.0000 - precision: 0.9680 - recall: 0.9477 - val_loss: 0.0074 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 176/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.1187e-04 - fn: 13.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 274.0000 - precision: 0.9682 - recall: 0.9547 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 177/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.3533e-04 - fn: 13.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 274.0000 - precision: 0.9716 - recall: 0.9547 - val_loss: 0.0076 - val_fn: 15.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 54.0000 - val_precision: 0.8438 - val_recall: 0.7826\n",
      "Epoch 178/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 4.5785e-04 - fn: 12.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 275.0000 - precision: 0.9649 - recall: 0.9582 - val_loss: 0.0073 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 179/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.8868e-04 - fn: 13.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 274.0000 - precision: 0.9856 - recall: 0.9547 - val_loss: 0.0073 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 180/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.5034e-04 - fn: 11.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 276.0000 - precision: 0.9753 - recall: 0.9617 - val_loss: 0.0073 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 181/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.5655e-04 - fn: 20.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 267.0000 - precision: 0.9709 - recall: 0.9303 - val_loss: 0.0073 - val_fn: 15.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 54.0000 - val_precision: 0.8438 - val_recall: 0.7826\n",
      "Epoch 182/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.8362e-04 - fn: 12.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 275.0000 - precision: 0.9752 - recall: 0.9582 - val_loss: 0.0074 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 183/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.4439e-04 - fn: 12.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 275.0000 - precision: 0.9786 - recall: 0.9582 - val_loss: 0.0073 - val_fn: 13.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 56.0000 - val_precision: 0.9180 - val_recall: 0.8116\n",
      "Epoch 184/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.8761e-04 - fn: 16.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 271.0000 - precision: 0.9713 - recall: 0.9443 - val_loss: 0.0075 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 185/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.0994e-04 - fn: 12.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 275.0000 - precision: 0.9649 - recall: 0.9582 - val_loss: 0.0072 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 186/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.4082e-04 - fn: 13.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 274.0000 - precision: 0.9751 - recall: 0.9547 - val_loss: 0.0071 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 187/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.1410e-04 - fn: 12.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 275.0000 - precision: 0.9752 - recall: 0.9582 - val_loss: 0.0078 - val_fn: 13.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 56.0000 - val_precision: 0.8889 - val_recall: 0.8116\n",
      "Epoch 188/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.5434e-04 - fn: 10.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 277.0000 - precision: 0.9754 - recall: 0.9652 - val_loss: 0.0071 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 189/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.3319e-04 - fn: 12.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 275.0000 - precision: 0.9683 - recall: 0.9582 - val_loss: 0.0069 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 190/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 4.3023e-04 - fn: 11.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 276.0000 - precision: 0.9787 - recall: 0.9617 - val_loss: 0.0089 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 191/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.8258e-04 - fn: 17.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 270.0000 - precision: 0.9747 - recall: 0.9408 - val_loss: 0.0086 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 192/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.4128e-04 - fn: 13.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 274.0000 - precision: 0.9751 - recall: 0.9547 - val_loss: 0.0090 - val_fn: 15.0000 - val_fp: 11.0000 - val_tn: 39793.0000 - val_tp: 54.0000 - val_precision: 0.8308 - val_recall: 0.7826\n",
      "Epoch 193/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.9385e-04 - fn: 16.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 271.0000 - precision: 0.9576 - recall: 0.9443 - val_loss: 0.0091 - val_fn: 13.0000 - val_fp: 11.0000 - val_tn: 39793.0000 - val_tp: 56.0000 - val_precision: 0.8358 - val_recall: 0.8116\n",
      "Epoch 194/300\n",
      "159491/159491 [==============================] - 1s 7us/sample - loss: 3.3682e-04 - fn: 11.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 276.0000 - precision: 0.9753 - recall: 0.9617 - val_loss: 0.0086 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 195/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.6446e-04 - fn: 13.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 274.0000 - precision: 0.9821 - recall: 0.9547 - val_loss: 0.0091 - val_fn: 13.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 56.0000 - val_precision: 0.8485 - val_recall: 0.8116\n",
      "Epoch 196/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.8366e-04 - fn: 15.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 272.0000 - precision: 0.9577 - recall: 0.9477 - val_loss: 0.0089 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 197/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.4827e-04 - fn: 14.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 273.0000 - precision: 0.9715 - recall: 0.9512 - val_loss: 0.0088 - val_fn: 16.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 53.0000 - val_precision: 0.8689 - val_recall: 0.7681\n",
      "Epoch 198/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.2030e-04 - fn: 11.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 276.0000 - precision: 0.9822 - recall: 0.9617 - val_loss: 0.0087 - val_fn: 14.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 55.0000 - val_precision: 0.8209 - val_recall: 0.7971\n",
      "Epoch 199/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.4667e-04 - fn: 11.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 276.0000 - precision: 0.9787 - recall: 0.9617 - val_loss: 0.0084 - val_fn: 14.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 55.0000 - val_precision: 0.8209 - val_recall: 0.7971\n",
      "Epoch 200/300\n",
      "159491/159491 [==============================] - 1s 7us/sample - loss: 2.4792e-04 - fn: 7.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 280.0000 - precision: 0.9894 - recall: 0.9756 - val_loss: 0.0083 - val_fn: 13.0000 - val_fp: 11.0000 - val_tn: 39793.0000 - val_tp: 56.0000 - val_precision: 0.8358 - val_recall: 0.8116\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 201/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 2.9593e-04 - fn: 12.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 275.0000 - precision: 0.9821 - recall: 0.9582 - val_loss: 0.0086 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 202/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.2765e-04 - fn: 12.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 275.0000 - precision: 0.9786 - recall: 0.9582 - val_loss: 0.0086 - val_fn: 14.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 55.0000 - val_precision: 0.8209 - val_recall: 0.7971\n",
      "Epoch 203/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.2758e-04 - fn: 9.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 278.0000 - precision: 0.9789 - recall: 0.9686 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 204/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.0022e-04 - fn: 10.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 277.0000 - precision: 0.9893 - recall: 0.9652 - val_loss: 0.0087 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 205/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.0212e-04 - fn: 10.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 277.0000 - precision: 0.9754 - recall: 0.9652 - val_loss: 0.0086 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 206/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.1498e-04 - fn: 8.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 279.0000 - precision: 0.9789 - recall: 0.9721 - val_loss: 0.0084 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 207/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.4476e-04 - fn: 9.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 278.0000 - precision: 0.9686 - recall: 0.9686 - val_loss: 0.0082 - val_fn: 15.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 54.0000 - val_precision: 0.8438 - val_recall: 0.7826\n",
      "Epoch 208/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.0360e-04 - fn: 12.0000 - fp: 11.0000 - tn: 159193.0000 - tp: 275.0000 - precision: 0.9615 - recall: 0.9582 - val_loss: 0.0083 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 209/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 2.1665e-04 - fn: 9.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 278.0000 - precision: 0.9823 - recall: 0.9686 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 210/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 5.7690e-04 - fn: 10.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 277.0000 - precision: 0.9754 - recall: 0.9652 - val_loss: 0.0079 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 211/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.2543e-04 - fn: 14.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 273.0000 - precision: 0.9681 - recall: 0.9512 - val_loss: 0.0081 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 212/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.7624e-04 - fn: 13.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 274.0000 - precision: 0.9751 - recall: 0.9547 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 213/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.8657e-04 - fn: 7.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 280.0000 - precision: 0.9825 - recall: 0.9756 - val_loss: 0.0079 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 214/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.7011e-04 - fn: 9.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 278.0000 - precision: 0.9653 - recall: 0.9686 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 215/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.5101e-04 - fn: 12.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 275.0000 - precision: 0.9717 - recall: 0.9582 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 216/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.3630e-04 - fn: 12.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 275.0000 - precision: 0.9752 - recall: 0.9582 - val_loss: 0.0081 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 217/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.2121e-04 - fn: 13.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 274.0000 - precision: 0.9648 - recall: 0.9547 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 218/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.1193e-04 - fn: 5.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 282.0000 - precision: 0.9826 - recall: 0.9826 - val_loss: 0.0083 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 219/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.7733e-04 - fn: 9.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 278.0000 - precision: 0.9858 - recall: 0.9686 - val_loss: 0.0084 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 220/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.5442e-04 - fn: 6.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 281.0000 - precision: 0.9825 - recall: 0.9791 - val_loss: 0.0079 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 221/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.3514e-04 - fn: 8.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 279.0000 - precision: 0.9859 - recall: 0.9721 - val_loss: 0.0080 - val_fn: 15.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 54.0000 - val_precision: 0.9000 - val_recall: 0.7826\n",
      "Epoch 222/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.5001e-04 - fn: 14.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 273.0000 - precision: 0.9715 - recall: 0.9512 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 223/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.6046e-04 - fn: 12.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 275.0000 - precision: 0.9752 - recall: 0.9582 - val_loss: 0.0081 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 224/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 4.6796e-04 - fn: 12.0000 - fp: 12.0000 - tn: 159192.0000 - tp: 275.0000 - precision: 0.9582 - recall: 0.9582 - val_loss: 0.0079 - val_fn: 13.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 56.0000 - val_precision: 0.8615 - val_recall: 0.8116\n",
      "Epoch 225/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 3.2884e-04 - fn: 12.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 275.0000 - precision: 0.9717 - recall: 0.9582 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 226/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.9344e-04 - fn: 8.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 279.0000 - precision: 0.9789 - recall: 0.9721 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 227/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.1822e-04 - fn: 10.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 277.0000 - precision: 0.9823 - recall: 0.9652 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 228/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.0098e-04 - fn: 9.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 278.0000 - precision: 0.9754 - recall: 0.9686 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 229/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.8460e-04 - fn: 9.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 278.0000 - precision: 0.9893 - recall: 0.9686 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 55.0000 - val_precision: 0.9167 - val_recall: 0.7971\n",
      "Epoch 230/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.7397e-04 - fn: 9.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 278.0000 - precision: 0.9720 - recall: 0.9686 - val_loss: 0.0078 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 231/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.0506e-04 - fn: 9.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 278.0000 - precision: 0.9720 - recall: 0.9686 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 232/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.5199e-04 - fn: 13.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 274.0000 - precision: 0.9682 - recall: 0.9547 - val_loss: 0.0078 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 233/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.1398e-04 - fn: 5.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 282.0000 - precision: 0.9826 - recall: 0.9826 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 234/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.1667e-04 - fn: 6.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 281.0000 - precision: 0.9860 - recall: 0.9791 - val_loss: 0.0078 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 235/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.2793e-04 - fn: 11.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 276.0000 - precision: 0.9892 - recall: 0.9617 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 236/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.8213e-04 - fn: 11.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 276.0000 - precision: 0.9650 - recall: 0.9617 - val_loss: 0.0074 - val_fn: 13.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 56.0000 - val_precision: 0.8750 - val_recall: 0.8116\n",
      "Epoch 237/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.4924e-04 - fn: 10.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 277.0000 - precision: 0.9788 - recall: 0.9652 - val_loss: 0.0074 - val_fn: 17.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 52.0000 - val_precision: 0.8814 - val_recall: 0.7536\n",
      "Epoch 238/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.1047e-04 - fn: 9.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 278.0000 - precision: 0.9858 - recall: 0.9686 - val_loss: 0.0075 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 239/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.2322e-04 - fn: 14.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 273.0000 - precision: 0.9647 - recall: 0.9512 - val_loss: 0.0073 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 240/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 2.9014e-04 - fn: 9.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 278.0000 - precision: 0.9789 - recall: 0.9686 - val_loss: 0.0074 - val_fn: 15.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 54.0000 - val_precision: 0.9000 - val_recall: 0.7826\n",
      "Epoch 241/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 1.9218e-04 - fn: 6.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 281.0000 - precision: 0.9860 - recall: 0.9791 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 242/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.6121e-04 - fn: 11.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 276.0000 - precision: 0.9787 - recall: 0.9617 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 243/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.7791e-04 - fn: 10.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 277.0000 - precision: 0.9788 - recall: 0.9652 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 244/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.3976e-04 - fn: 9.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 278.0000 - precision: 0.9754 - recall: 0.9686 - val_loss: 0.0080 - val_fn: 15.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 54.0000 - val_precision: 0.8438 - val_recall: 0.7826\n",
      "Epoch 245/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 3.0433e-04 - fn: 9.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 278.0000 - precision: 0.9823 - recall: 0.9686 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 246/300\n",
      "159491/159491 [==============================] - 1s 8us/sample - loss: 3.3722e-04 - fn: 9.0000 - fp: 10.0000 - tn: 159194.0000 - tp: 278.0000 - precision: 0.9653 - recall: 0.9686 - val_loss: 0.0078 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 247/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.2378e-04 - fn: 6.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 281.0000 - precision: 0.9791 - recall: 0.9791 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 55.0000 - val_precision: 0.9016 - val_recall: 0.7971\n",
      "Epoch 248/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.5690e-04 - fn: 8.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 279.0000 - precision: 0.9859 - recall: 0.9721 - val_loss: 0.0078 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 249/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.9655e-04 - fn: 8.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 279.0000 - precision: 0.9755 - recall: 0.9721 - val_loss: 0.0075 - val_fn: 16.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 53.0000 - val_precision: 0.8689 - val_recall: 0.7681\n",
      "Epoch 250/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 1.7893e-04 - fn: 8.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 279.0000 - precision: 0.9894 - recall: 0.9721 - val_loss: 0.0075 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 251/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.6015e-04 - fn: 9.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 278.0000 - precision: 0.9789 - recall: 0.9686 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 252/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.4657e-04 - fn: 8.0000 - fp: 9.0000 - tn: 159195.0000 - tp: 279.0000 - precision: 0.9688 - recall: 0.9721 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 55.0000 - val_precision: 0.9167 - val_recall: 0.7971\n",
      "Epoch 253/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.8365e-04 - fn: 14.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 273.0000 - precision: 0.9750 - recall: 0.9512 - val_loss: 0.0074 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 254/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 1.4369e-04 - fn: 3.0000 - fp: 2.0000 - tn: 159202.0000 - tp: 284.0000 - precision: 0.9930 - recall: 0.9895 - val_loss: 0.0075 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 255/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 2.8079e-04 - fn: 9.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 278.0000 - precision: 0.9720 - recall: 0.9686 - val_loss: 0.0075 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 256/300\n",
      "159491/159491 [==============================] - 1s 5us/sample - loss: 3.1924e-04 - fn: 9.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 278.0000 - precision: 0.9823 - recall: 0.9686 - val_loss: 0.0073 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 257/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 4.3393e-04 - fn: 12.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 275.0000 - precision: 0.9752 - recall: 0.9582 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 258/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.3417e-04 - fn: 8.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 279.0000 - precision: 0.9789 - recall: 0.9721 - val_loss: 0.0076 - val_fn: 15.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 54.0000 - val_precision: 0.8438 - val_recall: 0.7826\n",
      "Epoch 259/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 1.6980e-04 - fn: 6.0000 - fp: 2.0000 - tn: 159202.0000 - tp: 281.0000 - precision: 0.9929 - recall: 0.9791 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 260/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.1489e-04 - fn: 7.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 280.0000 - precision: 0.9825 - recall: 0.9756 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 261/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.6294e-04 - fn: 11.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 276.0000 - precision: 0.9753 - recall: 0.9617 - val_loss: 0.0076 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 262/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.6284e-04 - fn: 6.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 281.0000 - precision: 0.9757 - recall: 0.9791 - val_loss: 0.0075 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 263/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.4706e-04 - fn: 11.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 276.0000 - precision: 0.9857 - recall: 0.9617 - val_loss: 0.0076 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 264/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.0472e-04 - fn: 6.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 281.0000 - precision: 0.9757 - recall: 0.9791 - val_loss: 0.0075 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 265/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.2162e-04 - fn: 6.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 281.0000 - precision: 0.9860 - recall: 0.9791 - val_loss: 0.0074 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 266/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.2449e-04 - fn: 7.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 280.0000 - precision: 0.9790 - recall: 0.9756 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 267/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.4606e-04 - fn: 8.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 279.0000 - precision: 0.9755 - recall: 0.9721 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 268/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.0929e-04 - fn: 7.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 280.0000 - precision: 0.9859 - recall: 0.9756 - val_loss: 0.0078 - val_fn: 16.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 53.0000 - val_precision: 0.8833 - val_recall: 0.7681\n",
      "Epoch 269/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.5388e-04 - fn: 6.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 281.0000 - precision: 0.9894 - recall: 0.9791 - val_loss: 0.0076 - val_fn: 15.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 54.0000 - val_precision: 0.8852 - val_recall: 0.7826\n",
      "Epoch 270/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.1364e-04 - fn: 7.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 280.0000 - precision: 0.9894 - recall: 0.9756 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 271/300\n",
      "159491/159491 [==============================] - 1s 6us/sample - loss: 1.7150e-04 - fn: 5.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 282.0000 - precision: 0.9895 - recall: 0.9826 - val_loss: 0.0077 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 272/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.5569e-04 - fn: 12.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 275.0000 - precision: 0.9752 - recall: 0.9582 - val_loss: 0.0076 - val_fn: 13.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 56.0000 - val_precision: 0.9032 - val_recall: 0.8116\n",
      "Epoch 273/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.7193e-04 - fn: 12.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 275.0000 - precision: 0.9786 - recall: 0.9582 - val_loss: 0.0075 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 274/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 1.9626e-04 - fn: 6.0000 - fp: 2.0000 - tn: 159202.0000 - tp: 281.0000 - precision: 0.9929 - recall: 0.9791 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 275/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.4129e-04 - fn: 7.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 280.0000 - precision: 0.9722 - recall: 0.9756 - val_loss: 0.0082 - val_fn: 16.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 53.0000 - val_precision: 0.8154 - val_recall: 0.7681\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 276/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.4052e-04 - fn: 12.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 275.0000 - precision: 0.9717 - recall: 0.9582 - val_loss: 0.0080 - val_fn: 15.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 54.0000 - val_precision: 0.8438 - val_recall: 0.7826\n",
      "Epoch 277/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 1.8011e-04 - fn: 5.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 282.0000 - precision: 0.9895 - recall: 0.9826 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 278/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.3803e-04 - fn: 9.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 278.0000 - precision: 0.9720 - recall: 0.9686 - val_loss: 0.0081 - val_fn: 14.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 55.0000 - val_precision: 0.9167 - val_recall: 0.7971\n",
      "Epoch 279/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.8484e-04 - fn: 8.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 279.0000 - precision: 0.9824 - recall: 0.9721 - val_loss: 0.0082 - val_fn: 16.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 53.0000 - val_precision: 0.8548 - val_recall: 0.7681\n",
      "Epoch 280/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.3452e-04 - fn: 11.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 276.0000 - precision: 0.9857 - recall: 0.9617 - val_loss: 0.0081 - val_fn: 14.0000 - val_fp: 12.0000 - val_tn: 39792.0000 - val_tp: 55.0000 - val_precision: 0.8209 - val_recall: 0.7971\n",
      "Epoch 281/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.1094e-04 - fn: 5.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 282.0000 - precision: 0.9826 - recall: 0.9826 - val_loss: 0.0079 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 282/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.6310e-04 - fn: 10.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 277.0000 - precision: 0.9823 - recall: 0.9652 - val_loss: 0.0080 - val_fn: 14.0000 - val_fp: 11.0000 - val_tn: 39793.0000 - val_tp: 55.0000 - val_precision: 0.8333 - val_recall: 0.7971\n",
      "Epoch 283/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.0797e-04 - fn: 6.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 281.0000 - precision: 0.9825 - recall: 0.9791 - val_loss: 0.0080 - val_fn: 15.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 54.0000 - val_precision: 0.8710 - val_recall: 0.7826\n",
      "Epoch 284/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.2978e-04 - fn: 11.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 276.0000 - precision: 0.9857 - recall: 0.9617 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 285/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.2476e-04 - fn: 7.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 280.0000 - precision: 0.9859 - recall: 0.9756 - val_loss: 0.0079 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 286/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.3766e-04 - fn: 10.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 277.0000 - precision: 0.9893 - recall: 0.9652 - val_loss: 0.0083 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 287/300\n",
      "159491/159491 [==============================] - 1s 7us/sample - loss: 2.7314e-04 - fn: 8.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 279.0000 - precision: 0.9824 - recall: 0.9721 - val_loss: 0.0079 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 288/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.0733e-04 - fn: 8.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 279.0000 - precision: 0.9789 - recall: 0.9721 - val_loss: 0.0079 - val_fn: 14.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 55.0000 - val_precision: 0.9167 - val_recall: 0.7971\n",
      "Epoch 289/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.0631e-04 - fn: 8.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 279.0000 - precision: 0.9859 - recall: 0.9721 - val_loss: 0.0077 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n",
      "Epoch 290/300\n",
      "159491/159491 [==============================] - 1s 8us/sample - loss: 2.3255e-04 - fn: 10.0000 - fp: 7.0000 - tn: 159197.0000 - tp: 277.0000 - precision: 0.9754 - recall: 0.9652 - val_loss: 0.0079 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 291/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.1239e-04 - fn: 8.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 279.0000 - precision: 0.9859 - recall: 0.9721 - val_loss: 0.0080 - val_fn: 13.0000 - val_fp: 11.0000 - val_tn: 39793.0000 - val_tp: 56.0000 - val_precision: 0.8358 - val_recall: 0.8116\n",
      "Epoch 292/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.4092e-04 - fn: 7.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 280.0000 - precision: 0.9790 - recall: 0.9756 - val_loss: 0.0077 - val_fn: 15.0000 - val_fp: 5.0000 - val_tn: 39799.0000 - val_tp: 54.0000 - val_precision: 0.9153 - val_recall: 0.7826\n",
      "Epoch 293/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 1.9470e-04 - fn: 7.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 280.0000 - precision: 0.9859 - recall: 0.9756 - val_loss: 0.0079 - val_fn: 16.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 53.0000 - val_precision: 0.8833 - val_recall: 0.7681\n",
      "Epoch 294/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.0137e-04 - fn: 9.0000 - fp: 4.0000 - tn: 159200.0000 - tp: 278.0000 - precision: 0.9858 - recall: 0.9686 - val_loss: 0.0081 - val_fn: 15.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 54.0000 - val_precision: 0.8571 - val_recall: 0.7826\n",
      "Epoch 295/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.6342e-04 - fn: 10.0000 - fp: 5.0000 - tn: 159199.0000 - tp: 277.0000 - precision: 0.9823 - recall: 0.9652 - val_loss: 0.0083 - val_fn: 14.0000 - val_fp: 10.0000 - val_tn: 39794.0000 - val_tp: 55.0000 - val_precision: 0.8462 - val_recall: 0.7971\n",
      "Epoch 296/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.6914e-04 - fn: 11.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 276.0000 - precision: 0.9718 - recall: 0.9617 - val_loss: 0.0083 - val_fn: 14.0000 - val_fp: 8.0000 - val_tn: 39796.0000 - val_tp: 55.0000 - val_precision: 0.8730 - val_recall: 0.7971\n",
      "Epoch 297/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.7545e-04 - fn: 9.0000 - fp: 3.0000 - tn: 159201.0000 - tp: 278.0000 - precision: 0.9893 - recall: 0.9686 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 298/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 2.4927e-04 - fn: 7.0000 - fp: 8.0000 - tn: 159196.0000 - tp: 280.0000 - precision: 0.9722 - recall: 0.9756 - val_loss: 0.0082 - val_fn: 15.0000 - val_fp: 6.0000 - val_tn: 39798.0000 - val_tp: 54.0000 - val_precision: 0.9000 - val_recall: 0.7826\n",
      "Epoch 299/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 1.6695e-04 - fn: 6.0000 - fp: 2.0000 - tn: 159202.0000 - tp: 281.0000 - precision: 0.9929 - recall: 0.9791 - val_loss: 0.0082 - val_fn: 14.0000 - val_fp: 7.0000 - val_tn: 39797.0000 - val_tp: 55.0000 - val_precision: 0.8871 - val_recall: 0.7971\n",
      "Epoch 300/300\n",
      "159491/159491 [==============================] - 1s 4us/sample - loss: 3.2799e-04 - fn: 9.0000 - fp: 6.0000 - tn: 159198.0000 - tp: 278.0000 - precision: 0.9789 - recall: 0.9686 - val_loss: 0.0081 - val_fn: 14.0000 - val_fp: 9.0000 - val_tn: 39795.0000 - val_tp: 55.0000 - val_precision: 0.8594 - val_recall: 0.7971\n"
     ]
    }
   ],
   "source": [
    "METRICS = [\n",
    "#     keras.metrics.Accuracy(name='accuracy'),\n",
    "    keras.metrics.FalseNegatives(name='fn'),\n",
    "    keras.metrics.FalsePositives(name='fp'),\n",
    "    keras.metrics.TrueNegatives(name='tn'),\n",
    "    keras.metrics.TruePositives(name='tp'),\n",
    "    keras.metrics.Precision(name='precision'),\n",
    "    keras.metrics.Recall(name='recall')\n",
    "]\n",
    "\n",
    "model.compile(optimizer=keras.optimizers.Adam(1e-3), loss='binary_crossentropy', metrics=METRICS)\n",
    "\n",
    "callbacks = [keras.callbacks.ModelCheckpoint('fraud_model_at_epoch_{epoch}.h5')]\n",
    "class_weight = {0:w_p, 1:w_n}\n",
    "\n",
    "r = model.fit(\n",
    "    X_train, y_train, \n",
    "    validation_data=(X_validate, y_validate),\n",
    "    batch_size=2048, \n",
    "    epochs=300, \n",
    "#     class_weight=class_weight,\n",
    "    callbacks=callbacks,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "85443/85443 [==============================] - 10s 122us/sample - loss: 0.0047 - fn: 26.0000 - fp: 16.0000 - tn: 85291.0000 - tp: 110.0000 - precision: 0.8730 - recall: 0.8088\n",
      "[0.004723706854455897, 26.0, 16.0, 85291.0, 110.0, 0.8730159, 0.8088235]\n"
     ]
    }
   ],
   "source": [
    "score = model.evaluate(X_test, y_test)\n",
    "print(score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fee27cd9dd8>"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x1152 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 16))\n",
    "\n",
    "plt.subplot(4, 2, 1)\n",
    "plt.plot(r.history['loss'], label='Loss')\n",
    "plt.plot(r.history['val_loss'], label='val_Loss')\n",
    "plt.title('Loss Function evolution during training')\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(4, 2, 2)\n",
    "plt.plot(r.history['fn'], label='fn')\n",
    "plt.plot(r.history['val_fn'], label='val_fn')\n",
    "plt.title('Accuracy evolution during training')\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(4, 2, 3)\n",
    "plt.plot(r.history['precision'], label='precision')\n",
    "plt.plot(r.history['val_precision'], label='val_precision')\n",
    "plt.title('Precision evolution during training')\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(4, 2, 4)\n",
    "plt.plot(r.history['recall'], label='recall')\n",
    "plt.plot(r.history['val_recall'], label='val_recall')\n",
    "plt.title('Recall evolution during training')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Result:\n",
      "================================================\n",
      "Accuracy Score: 100.00%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                  0      1  accuracy  macro avg  weighted avg\n",
      "precision      1.00   0.99      1.00       1.00          1.00\n",
      "recall         1.00   1.00      1.00       1.00          1.00\n",
      "f1-score       1.00   0.99      1.00       1.00          1.00\n",
      "support   159204.00 287.00      1.00  159491.00     159491.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[159202      2]\n",
      " [     1    286]]\n",
      "\n",
      "Test Result:\n",
      "================================================\n",
      "Accuracy Score: 99.95%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                 0      1  accuracy  macro avg  weighted avg\n",
      "precision     1.00   0.87      1.00       0.94          1.00\n",
      "recall        1.00   0.81      1.00       0.90          1.00\n",
      "f1-score      1.00   0.84      1.00       0.92          1.00\n",
      "support   85307.00 136.00      1.00   85443.00      85443.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[85291    16]\n",
      " [   26   110]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "y_train_pred = model.predict(X_train)\n",
    "y_test_pred = model.predict(X_test)\n",
    "\n",
    "print_score(y_train, y_train_pred.round(), train=True)\n",
    "print_score(y_test, y_test_pred.round(), train=False)\n",
    "\n",
    "scores_dict = {\n",
    "    'ANNs': {\n",
    "        'Train': f1_score(y_train, y_train_pred.round()),\n",
    "        'Test': f1_score(y_test, y_test_pred.round()),\n",
    "    },\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. 2. XGBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Result:\n",
      "================================================\n",
      "Accuracy Score: 99.97%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                  0      1  accuracy  macro avg  weighted avg\n",
      "precision      1.00   0.98      1.00       0.99          1.00\n",
      "recall         1.00   0.84      1.00       0.92          1.00\n",
      "f1-score       1.00   0.91      1.00       0.95          1.00\n",
      "support   159204.00 287.00      1.00  159491.00     159491.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[159200      4]\n",
      " [    46    241]]\n",
      "\n",
      "Test Result:\n",
      "================================================\n",
      "Accuracy Score: 99.96%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                 0      1  accuracy  macro avg  weighted avg\n",
      "precision     1.00   0.93      1.00       0.97          1.00\n",
      "recall        1.00   0.82      1.00       0.91          1.00\n",
      "f1-score      1.00   0.87      1.00       0.94          1.00\n",
      "support   85307.00 136.00      1.00   85443.00      85443.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[85299     8]\n",
      " [   25   111]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from xgboost import XGBClassifier\n",
    "\n",
    "xgb_clf = XGBClassifier()\n",
    "xgb_clf.fit(X_train, y_train, eval_metric='aucpr')\n",
    "\n",
    "y_train_pred = xgb_clf.predict(X_train)\n",
    "y_test_pred = xgb_clf.predict(X_test)\n",
    "\n",
    "print_score(y_train, y_train_pred, train=True)\n",
    "print_score(y_test, y_test_pred, train=False)\n",
    "\n",
    "scores_dict['XGBoost'] = {\n",
    "        'Train': f1_score(y_train,y_train_pred),\n",
    "        'Test': f1_score(y_test, y_test_pred),\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. 3. Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Result:\n",
      "================================================\n",
      "Accuracy Score: 100.00%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                  0      1  accuracy  macro avg  weighted avg\n",
      "precision      1.00   1.00      1.00       1.00          1.00\n",
      "recall         1.00   1.00      1.00       1.00          1.00\n",
      "f1-score       1.00   1.00      1.00       1.00          1.00\n",
      "support   159204.00 287.00      1.00  159491.00     159491.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[159204      0]\n",
      " [     1    286]]\n",
      "\n",
      "Test Result:\n",
      "================================================\n",
      "Accuracy Score: 99.96%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                 0      1  accuracy  macro avg  weighted avg\n",
      "precision     1.00   0.93      1.00       0.96          1.00\n",
      "recall        1.00   0.82      1.00       0.91          1.00\n",
      "f1-score      1.00   0.87      1.00       0.93          1.00\n",
      "support   85307.00 136.00      1.00   85443.00      85443.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[85298     9]\n",
      " [   25   111]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "rf_clf = RandomForestClassifier(n_estimators=100, oob_score=False)\n",
    "rf_clf.fit(X_train, y_train)\n",
    "\n",
    "y_train_pred = rf_clf.predict(X_train)\n",
    "y_test_pred = rf_clf.predict(X_test)\n",
    "\n",
    "print_score(y_train, y_train_pred, train=True)\n",
    "print_score(y_test, y_test_pred, train=False)\n",
    "\n",
    "scores_dict['Random Forest'] = {\n",
    "        'Train': f1_score(y_train,y_train_pred),\n",
    "        'Test': f1_score(y_test, y_test_pred),\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. 4. CatBoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "_kg_hide-output": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Learning rate set to 0.070388\n",
      "0:\tlearn: 0.4597542\ttotal: 73.3ms\tremaining: 1m 13s\n",
      "1:\tlearn: 0.3129054\ttotal: 147ms\tremaining: 1m 13s\n",
      "2:\tlearn: 0.2124019\ttotal: 223ms\tremaining: 1m 14s\n",
      "3:\tlearn: 0.1425771\ttotal: 294ms\tremaining: 1m 13s\n",
      "4:\tlearn: 0.0941945\ttotal: 365ms\tremaining: 1m 12s\n",
      "5:\tlearn: 0.0642168\ttotal: 436ms\tremaining: 1m 12s\n",
      "6:\tlearn: 0.0445214\ttotal: 508ms\tremaining: 1m 12s\n",
      "7:\tlearn: 0.0320400\ttotal: 575ms\tremaining: 1m 11s\n",
      "8:\tlearn: 0.0237219\ttotal: 644ms\tremaining: 1m 10s\n",
      "9:\tlearn: 0.0180726\ttotal: 709ms\tremaining: 1m 10s\n",
      "10:\tlearn: 0.0142287\ttotal: 778ms\tremaining: 1m 9s\n",
      "11:\tlearn: 0.0114157\ttotal: 848ms\tremaining: 1m 9s\n",
      "12:\tlearn: 0.0094174\ttotal: 913ms\tremaining: 1m 9s\n",
      "13:\tlearn: 0.0079424\ttotal: 980ms\tremaining: 1m 9s\n",
      "14:\tlearn: 0.0069280\ttotal: 1.05s\tremaining: 1m 8s\n",
      "15:\tlearn: 0.0060589\ttotal: 1.12s\tremaining: 1m 8s\n",
      "16:\tlearn: 0.0053785\ttotal: 1.19s\tremaining: 1m 8s\n",
      "17:\tlearn: 0.0049026\ttotal: 1.26s\tremaining: 1m 8s\n",
      "18:\tlearn: 0.0045010\ttotal: 1.33s\tremaining: 1m 8s\n",
      "19:\tlearn: 0.0042153\ttotal: 1.44s\tremaining: 1m 10s\n",
      "20:\tlearn: 0.0039452\ttotal: 1.61s\tremaining: 1m 15s\n",
      "21:\tlearn: 0.0037028\ttotal: 1.73s\tremaining: 1m 17s\n",
      "22:\tlearn: 0.0035388\ttotal: 1.88s\tremaining: 1m 20s\n",
      "23:\tlearn: 0.0033875\ttotal: 2.02s\tremaining: 1m 22s\n",
      "24:\tlearn: 0.0032854\ttotal: 2.13s\tremaining: 1m 23s\n",
      "25:\tlearn: 0.0032028\ttotal: 2.21s\tremaining: 1m 22s\n",
      "26:\tlearn: 0.0031281\ttotal: 2.28s\tremaining: 1m 22s\n",
      "27:\tlearn: 0.0030479\ttotal: 2.35s\tremaining: 1m 21s\n",
      "28:\tlearn: 0.0029862\ttotal: 2.42s\tremaining: 1m 20s\n",
      "29:\tlearn: 0.0029433\ttotal: 2.48s\tremaining: 1m 20s\n",
      "30:\tlearn: 0.0028923\ttotal: 2.56s\tremaining: 1m 19s\n",
      "31:\tlearn: 0.0028525\ttotal: 2.63s\tremaining: 1m 19s\n",
      "32:\tlearn: 0.0028196\ttotal: 2.69s\tremaining: 1m 18s\n",
      "33:\tlearn: 0.0027896\ttotal: 2.77s\tremaining: 1m 18s\n",
      "34:\tlearn: 0.0027599\ttotal: 2.83s\tremaining: 1m 18s\n",
      "35:\tlearn: 0.0027341\ttotal: 2.9s\tremaining: 1m 17s\n",
      "36:\tlearn: 0.0027177\ttotal: 2.97s\tremaining: 1m 17s\n",
      "37:\tlearn: 0.0026748\ttotal: 3.04s\tremaining: 1m 16s\n",
      "38:\tlearn: 0.0026516\ttotal: 3.11s\tremaining: 1m 16s\n",
      "39:\tlearn: 0.0025952\ttotal: 3.18s\tremaining: 1m 16s\n",
      "40:\tlearn: 0.0025593\ttotal: 3.25s\tremaining: 1m 16s\n",
      "41:\tlearn: 0.0025233\ttotal: 3.32s\tremaining: 1m 15s\n",
      "42:\tlearn: 0.0024811\ttotal: 3.4s\tremaining: 1m 15s\n",
      "43:\tlearn: 0.0024567\ttotal: 3.46s\tremaining: 1m 15s\n",
      "44:\tlearn: 0.0024363\ttotal: 3.53s\tremaining: 1m 14s\n",
      "45:\tlearn: 0.0024171\ttotal: 3.6s\tremaining: 1m 14s\n",
      "46:\tlearn: 0.0023981\ttotal: 3.67s\tremaining: 1m 14s\n",
      "47:\tlearn: 0.0023650\ttotal: 3.74s\tremaining: 1m 14s\n",
      "48:\tlearn: 0.0023445\ttotal: 3.81s\tremaining: 1m 13s\n",
      "49:\tlearn: 0.0023278\ttotal: 3.88s\tremaining: 1m 13s\n",
      "50:\tlearn: 0.0022982\ttotal: 3.95s\tremaining: 1m 13s\n",
      "51:\tlearn: 0.0022708\ttotal: 4.02s\tremaining: 1m 13s\n",
      "52:\tlearn: 0.0022497\ttotal: 4.09s\tremaining: 1m 13s\n",
      "53:\tlearn: 0.0022405\ttotal: 4.15s\tremaining: 1m 12s\n",
      "54:\tlearn: 0.0022282\ttotal: 4.22s\tremaining: 1m 12s\n",
      "55:\tlearn: 0.0022208\ttotal: 4.29s\tremaining: 1m 12s\n",
      "56:\tlearn: 0.0022074\ttotal: 4.36s\tremaining: 1m 12s\n",
      "57:\tlearn: 0.0021777\ttotal: 4.43s\tremaining: 1m 11s\n",
      "58:\tlearn: 0.0021554\ttotal: 4.5s\tremaining: 1m 11s\n",
      "59:\tlearn: 0.0021393\ttotal: 4.57s\tremaining: 1m 11s\n",
      "60:\tlearn: 0.0021296\ttotal: 4.64s\tremaining: 1m 11s\n",
      "61:\tlearn: 0.0021170\ttotal: 4.71s\tremaining: 1m 11s\n",
      "62:\tlearn: 0.0020998\ttotal: 4.78s\tremaining: 1m 11s\n",
      "63:\tlearn: 0.0020898\ttotal: 4.84s\tremaining: 1m 10s\n",
      "64:\tlearn: 0.0020789\ttotal: 4.92s\tremaining: 1m 10s\n",
      "65:\tlearn: 0.0020699\ttotal: 4.99s\tremaining: 1m 10s\n",
      "66:\tlearn: 0.0020625\ttotal: 5.05s\tremaining: 1m 10s\n",
      "67:\tlearn: 0.0020551\ttotal: 5.12s\tremaining: 1m 10s\n",
      "68:\tlearn: 0.0020453\ttotal: 5.19s\tremaining: 1m 10s\n",
      "69:\tlearn: 0.0020366\ttotal: 5.26s\tremaining: 1m 9s\n",
      "70:\tlearn: 0.0020222\ttotal: 5.33s\tremaining: 1m 9s\n",
      "71:\tlearn: 0.0020082\ttotal: 5.41s\tremaining: 1m 9s\n",
      "72:\tlearn: 0.0019954\ttotal: 5.48s\tremaining: 1m 9s\n",
      "73:\tlearn: 0.0019860\ttotal: 5.55s\tremaining: 1m 9s\n",
      "74:\tlearn: 0.0019801\ttotal: 5.62s\tremaining: 1m 9s\n",
      "75:\tlearn: 0.0019671\ttotal: 5.69s\tremaining: 1m 9s\n",
      "76:\tlearn: 0.0019544\ttotal: 5.76s\tremaining: 1m 9s\n",
      "77:\tlearn: 0.0019409\ttotal: 5.83s\tremaining: 1m 8s\n",
      "78:\tlearn: 0.0019358\ttotal: 5.9s\tremaining: 1m 8s\n",
      "79:\tlearn: 0.0019313\ttotal: 5.97s\tremaining: 1m 8s\n",
      "80:\tlearn: 0.0019256\ttotal: 6.04s\tremaining: 1m 8s\n",
      "81:\tlearn: 0.0019179\ttotal: 6.1s\tremaining: 1m 8s\n",
      "82:\tlearn: 0.0019136\ttotal: 6.18s\tremaining: 1m 8s\n",
      "83:\tlearn: 0.0019091\ttotal: 6.24s\tremaining: 1m 8s\n",
      "84:\tlearn: 0.0018982\ttotal: 6.31s\tremaining: 1m 7s\n",
      "85:\tlearn: 0.0018944\ttotal: 6.38s\tremaining: 1m 7s\n",
      "86:\tlearn: 0.0018811\ttotal: 6.45s\tremaining: 1m 7s\n",
      "87:\tlearn: 0.0018722\ttotal: 6.52s\tremaining: 1m 7s\n",
      "88:\tlearn: 0.0018650\ttotal: 6.59s\tremaining: 1m 7s\n",
      "89:\tlearn: 0.0018567\ttotal: 6.66s\tremaining: 1m 7s\n",
      "90:\tlearn: 0.0018532\ttotal: 6.72s\tremaining: 1m 7s\n",
      "91:\tlearn: 0.0018442\ttotal: 6.79s\tremaining: 1m 7s\n",
      "92:\tlearn: 0.0018376\ttotal: 6.86s\tremaining: 1m 6s\n",
      "93:\tlearn: 0.0018339\ttotal: 6.93s\tremaining: 1m 6s\n",
      "94:\tlearn: 0.0018271\ttotal: 7s\tremaining: 1m 6s\n",
      "95:\tlearn: 0.0018178\ttotal: 7.07s\tremaining: 1m 6s\n",
      "96:\tlearn: 0.0017978\ttotal: 7.14s\tremaining: 1m 6s\n",
      "97:\tlearn: 0.0017906\ttotal: 7.21s\tremaining: 1m 6s\n",
      "98:\tlearn: 0.0017777\ttotal: 7.29s\tremaining: 1m 6s\n",
      "99:\tlearn: 0.0017731\ttotal: 7.36s\tremaining: 1m 6s\n",
      "100:\tlearn: 0.0017604\ttotal: 7.43s\tremaining: 1m 6s\n",
      "101:\tlearn: 0.0017552\ttotal: 7.5s\tremaining: 1m 5s\n",
      "102:\tlearn: 0.0017418\ttotal: 7.56s\tremaining: 1m 5s\n",
      "103:\tlearn: 0.0017367\ttotal: 7.64s\tremaining: 1m 5s\n",
      "104:\tlearn: 0.0017330\ttotal: 7.7s\tremaining: 1m 5s\n",
      "105:\tlearn: 0.0017284\ttotal: 7.79s\tremaining: 1m 5s\n",
      "106:\tlearn: 0.0017247\ttotal: 7.86s\tremaining: 1m 5s\n",
      "107:\tlearn: 0.0017209\ttotal: 7.93s\tremaining: 1m 5s\n",
      "108:\tlearn: 0.0017151\ttotal: 7.99s\tremaining: 1m 5s\n",
      "109:\tlearn: 0.0017070\ttotal: 8.06s\tremaining: 1m 5s\n",
      "110:\tlearn: 0.0017013\ttotal: 8.13s\tremaining: 1m 5s\n",
      "111:\tlearn: 0.0016970\ttotal: 8.19s\tremaining: 1m 4s\n",
      "112:\tlearn: 0.0016882\ttotal: 8.26s\tremaining: 1m 4s\n",
      "113:\tlearn: 0.0016839\ttotal: 8.34s\tremaining: 1m 4s\n",
      "114:\tlearn: 0.0016812\ttotal: 8.4s\tremaining: 1m 4s\n",
      "115:\tlearn: 0.0016720\ttotal: 8.47s\tremaining: 1m 4s\n",
      "116:\tlearn: 0.0016641\ttotal: 8.54s\tremaining: 1m 4s\n",
      "117:\tlearn: 0.0016537\ttotal: 8.62s\tremaining: 1m 4s\n",
      "118:\tlearn: 0.0016482\ttotal: 8.68s\tremaining: 1m 4s\n",
      "119:\tlearn: 0.0016425\ttotal: 8.76s\tremaining: 1m 4s\n",
      "120:\tlearn: 0.0016363\ttotal: 8.83s\tremaining: 1m 4s\n",
      "121:\tlearn: 0.0016294\ttotal: 8.89s\tremaining: 1m 4s\n",
      "122:\tlearn: 0.0016266\ttotal: 8.96s\tremaining: 1m 3s\n",
      "123:\tlearn: 0.0016176\ttotal: 9.04s\tremaining: 1m 3s\n",
      "124:\tlearn: 0.0016118\ttotal: 9.1s\tremaining: 1m 3s\n",
      "125:\tlearn: 0.0016022\ttotal: 9.18s\tremaining: 1m 3s\n",
      "126:\tlearn: 0.0015976\ttotal: 9.24s\tremaining: 1m 3s\n",
      "127:\tlearn: 0.0015913\ttotal: 9.31s\tremaining: 1m 3s\n",
      "128:\tlearn: 0.0015877\ttotal: 9.38s\tremaining: 1m 3s\n",
      "129:\tlearn: 0.0015801\ttotal: 9.45s\tremaining: 1m 3s\n",
      "130:\tlearn: 0.0015756\ttotal: 9.52s\tremaining: 1m 3s\n",
      "131:\tlearn: 0.0015694\ttotal: 9.59s\tremaining: 1m 3s\n",
      "132:\tlearn: 0.0015639\ttotal: 9.66s\tremaining: 1m 2s\n",
      "133:\tlearn: 0.0015530\ttotal: 9.73s\tremaining: 1m 2s\n",
      "134:\tlearn: 0.0015447\ttotal: 9.8s\tremaining: 1m 2s\n",
      "135:\tlearn: 0.0015417\ttotal: 9.87s\tremaining: 1m 2s\n",
      "136:\tlearn: 0.0015385\ttotal: 9.94s\tremaining: 1m 2s\n",
      "137:\tlearn: 0.0015355\ttotal: 10s\tremaining: 1m 2s\n",
      "138:\tlearn: 0.0015309\ttotal: 10.1s\tremaining: 1m 2s\n",
      "139:\tlearn: 0.0015279\ttotal: 10.2s\tremaining: 1m 2s\n",
      "140:\tlearn: 0.0015154\ttotal: 10.2s\tremaining: 1m 2s\n",
      "141:\tlearn: 0.0015125\ttotal: 10.3s\tremaining: 1m 2s\n",
      "142:\tlearn: 0.0015094\ttotal: 10.5s\tremaining: 1m 2s\n",
      "143:\tlearn: 0.0015065\ttotal: 10.6s\tremaining: 1m 2s\n",
      "144:\tlearn: 0.0015043\ttotal: 10.7s\tremaining: 1m 3s\n",
      "145:\tlearn: 0.0014979\ttotal: 10.8s\tremaining: 1m 3s\n",
      "146:\tlearn: 0.0014931\ttotal: 11s\tremaining: 1m 3s\n",
      "147:\tlearn: 0.0014900\ttotal: 11.1s\tremaining: 1m 3s\n",
      "148:\tlearn: 0.0014833\ttotal: 11.2s\tremaining: 1m 4s\n",
      "149:\tlearn: 0.0014724\ttotal: 11.4s\tremaining: 1m 4s\n",
      "150:\tlearn: 0.0014692\ttotal: 11.5s\tremaining: 1m 4s\n",
      "151:\tlearn: 0.0014602\ttotal: 11.6s\tremaining: 1m 4s\n",
      "152:\tlearn: 0.0014481\ttotal: 11.7s\tremaining: 1m 4s\n",
      "153:\tlearn: 0.0014425\ttotal: 11.7s\tremaining: 1m 4s\n",
      "154:\tlearn: 0.0014382\ttotal: 11.8s\tremaining: 1m 4s\n",
      "155:\tlearn: 0.0014299\ttotal: 11.9s\tremaining: 1m 4s\n",
      "156:\tlearn: 0.0014228\ttotal: 12s\tremaining: 1m 4s\n",
      "157:\tlearn: 0.0014200\ttotal: 12s\tremaining: 1m 4s\n",
      "158:\tlearn: 0.0014173\ttotal: 12.1s\tremaining: 1m 4s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "159:\tlearn: 0.0014158\ttotal: 12.2s\tremaining: 1m 3s\n",
      "160:\tlearn: 0.0014116\ttotal: 12.2s\tremaining: 1m 3s\n",
      "161:\tlearn: 0.0014042\ttotal: 12.3s\tremaining: 1m 3s\n",
      "162:\tlearn: 0.0014008\ttotal: 12.4s\tremaining: 1m 3s\n",
      "163:\tlearn: 0.0013983\ttotal: 12.6s\tremaining: 1m 4s\n",
      "164:\tlearn: 0.0013943\ttotal: 12.7s\tremaining: 1m 4s\n",
      "165:\tlearn: 0.0013925\ttotal: 12.8s\tremaining: 1m 4s\n",
      "166:\tlearn: 0.0013844\ttotal: 13s\tremaining: 1m 4s\n",
      "167:\tlearn: 0.0013819\ttotal: 13.1s\tremaining: 1m 4s\n",
      "168:\tlearn: 0.0013784\ttotal: 13.2s\tremaining: 1m 4s\n",
      "169:\tlearn: 0.0013745\ttotal: 13.3s\tremaining: 1m 4s\n",
      "170:\tlearn: 0.0013731\ttotal: 13.3s\tremaining: 1m 4s\n",
      "171:\tlearn: 0.0013626\ttotal: 13.4s\tremaining: 1m 4s\n",
      "172:\tlearn: 0.0013588\ttotal: 13.5s\tremaining: 1m 4s\n",
      "173:\tlearn: 0.0013574\ttotal: 13.6s\tremaining: 1m 4s\n",
      "174:\tlearn: 0.0013531\ttotal: 13.6s\tremaining: 1m 4s\n",
      "175:\tlearn: 0.0013475\ttotal: 13.7s\tremaining: 1m 4s\n",
      "176:\tlearn: 0.0013445\ttotal: 13.8s\tremaining: 1m 4s\n",
      "177:\tlearn: 0.0013404\ttotal: 13.9s\tremaining: 1m 3s\n",
      "178:\tlearn: 0.0013367\ttotal: 13.9s\tremaining: 1m 3s\n",
      "179:\tlearn: 0.0013341\ttotal: 14s\tremaining: 1m 3s\n",
      "180:\tlearn: 0.0013267\ttotal: 14.1s\tremaining: 1m 3s\n",
      "181:\tlearn: 0.0013187\ttotal: 14.1s\tremaining: 1m 3s\n",
      "182:\tlearn: 0.0013056\ttotal: 14.2s\tremaining: 1m 3s\n",
      "183:\tlearn: 0.0013026\ttotal: 14.3s\tremaining: 1m 3s\n",
      "184:\tlearn: 0.0012973\ttotal: 14.3s\tremaining: 1m 3s\n",
      "185:\tlearn: 0.0012877\ttotal: 14.4s\tremaining: 1m 3s\n",
      "186:\tlearn: 0.0012851\ttotal: 14.5s\tremaining: 1m 2s\n",
      "187:\tlearn: 0.0012815\ttotal: 14.6s\tremaining: 1m 2s\n",
      "188:\tlearn: 0.0012785\ttotal: 14.6s\tremaining: 1m 2s\n",
      "189:\tlearn: 0.0012764\ttotal: 14.7s\tremaining: 1m 2s\n",
      "190:\tlearn: 0.0012740\ttotal: 14.8s\tremaining: 1m 2s\n",
      "191:\tlearn: 0.0012690\ttotal: 14.8s\tremaining: 1m 2s\n",
      "192:\tlearn: 0.0012668\ttotal: 14.9s\tremaining: 1m 2s\n",
      "193:\tlearn: 0.0012646\ttotal: 15s\tremaining: 1m 2s\n",
      "194:\tlearn: 0.0012623\ttotal: 15s\tremaining: 1m 2s\n",
      "195:\tlearn: 0.0012609\ttotal: 15.1s\tremaining: 1m 2s\n",
      "196:\tlearn: 0.0012587\ttotal: 15.2s\tremaining: 1m 1s\n",
      "197:\tlearn: 0.0012563\ttotal: 15.3s\tremaining: 1m 1s\n",
      "198:\tlearn: 0.0012543\ttotal: 15.3s\tremaining: 1m 1s\n",
      "199:\tlearn: 0.0012483\ttotal: 15.4s\tremaining: 1m 1s\n",
      "200:\tlearn: 0.0012468\ttotal: 15.5s\tremaining: 1m 1s\n",
      "201:\tlearn: 0.0012452\ttotal: 15.6s\tremaining: 1m 1s\n",
      "202:\tlearn: 0.0012377\ttotal: 15.6s\tremaining: 1m 1s\n",
      "203:\tlearn: 0.0012360\ttotal: 15.7s\tremaining: 1m 1s\n",
      "204:\tlearn: 0.0012317\ttotal: 15.8s\tremaining: 1m 1s\n",
      "205:\tlearn: 0.0012300\ttotal: 15.9s\tremaining: 1m 1s\n",
      "206:\tlearn: 0.0012286\ttotal: 15.9s\tremaining: 1m\n",
      "207:\tlearn: 0.0012263\ttotal: 16s\tremaining: 1m\n",
      "208:\tlearn: 0.0012251\ttotal: 16.1s\tremaining: 1m\n",
      "209:\tlearn: 0.0012235\ttotal: 16.1s\tremaining: 1m\n",
      "210:\tlearn: 0.0012179\ttotal: 16.2s\tremaining: 1m\n",
      "211:\tlearn: 0.0012102\ttotal: 16.3s\tremaining: 1m\n",
      "212:\tlearn: 0.0012074\ttotal: 16.3s\tremaining: 1m\n",
      "213:\tlearn: 0.0012053\ttotal: 16.4s\tremaining: 1m\n",
      "214:\tlearn: 0.0011995\ttotal: 16.5s\tremaining: 1m\n",
      "215:\tlearn: 0.0011952\ttotal: 16.5s\tremaining: 1m\n",
      "216:\tlearn: 0.0011902\ttotal: 16.6s\tremaining: 59.9s\n",
      "217:\tlearn: 0.0011825\ttotal: 16.7s\tremaining: 59.8s\n",
      "218:\tlearn: 0.0011805\ttotal: 16.7s\tremaining: 59.7s\n",
      "219:\tlearn: 0.0011799\ttotal: 16.8s\tremaining: 59.6s\n",
      "220:\tlearn: 0.0011782\ttotal: 16.9s\tremaining: 59.5s\n",
      "221:\tlearn: 0.0011765\ttotal: 17s\tremaining: 59.4s\n",
      "222:\tlearn: 0.0011694\ttotal: 17s\tremaining: 59.3s\n",
      "223:\tlearn: 0.0011658\ttotal: 17.1s\tremaining: 59.2s\n",
      "224:\tlearn: 0.0011629\ttotal: 17.2s\tremaining: 59.1s\n",
      "225:\tlearn: 0.0011566\ttotal: 17.2s\tremaining: 59s\n",
      "226:\tlearn: 0.0011550\ttotal: 17.3s\tremaining: 58.9s\n",
      "227:\tlearn: 0.0011486\ttotal: 17.4s\tremaining: 58.8s\n",
      "228:\tlearn: 0.0011477\ttotal: 17.4s\tremaining: 58.7s\n",
      "229:\tlearn: 0.0011457\ttotal: 17.5s\tremaining: 58.6s\n",
      "230:\tlearn: 0.0011440\ttotal: 17.6s\tremaining: 58.5s\n",
      "231:\tlearn: 0.0011422\ttotal: 17.7s\tremaining: 58.4s\n",
      "232:\tlearn: 0.0011381\ttotal: 17.7s\tremaining: 58.3s\n",
      "233:\tlearn: 0.0011365\ttotal: 17.8s\tremaining: 58.2s\n",
      "234:\tlearn: 0.0011346\ttotal: 17.9s\tremaining: 58.1s\n",
      "235:\tlearn: 0.0011337\ttotal: 17.9s\tremaining: 58.1s\n",
      "236:\tlearn: 0.0011299\ttotal: 18s\tremaining: 58s\n",
      "237:\tlearn: 0.0011229\ttotal: 18.1s\tremaining: 57.9s\n",
      "238:\tlearn: 0.0011176\ttotal: 18.2s\tremaining: 57.8s\n",
      "239:\tlearn: 0.0011150\ttotal: 18.2s\tremaining: 57.7s\n",
      "240:\tlearn: 0.0011116\ttotal: 18.3s\tremaining: 57.6s\n",
      "241:\tlearn: 0.0011108\ttotal: 18.4s\tremaining: 57.5s\n",
      "242:\tlearn: 0.0011087\ttotal: 18.4s\tremaining: 57.4s\n",
      "243:\tlearn: 0.0011066\ttotal: 18.5s\tremaining: 57.3s\n",
      "244:\tlearn: 0.0011060\ttotal: 18.6s\tremaining: 57.2s\n",
      "245:\tlearn: 0.0011043\ttotal: 18.6s\tremaining: 57.1s\n",
      "246:\tlearn: 0.0011002\ttotal: 18.7s\tremaining: 57.1s\n",
      "247:\tlearn: 0.0010995\ttotal: 18.8s\tremaining: 57s\n",
      "248:\tlearn: 0.0010965\ttotal: 18.9s\tremaining: 56.9s\n",
      "249:\tlearn: 0.0010942\ttotal: 18.9s\tremaining: 56.8s\n",
      "250:\tlearn: 0.0010922\ttotal: 19s\tremaining: 56.7s\n",
      "251:\tlearn: 0.0010877\ttotal: 19.1s\tremaining: 56.6s\n",
      "252:\tlearn: 0.0010839\ttotal: 19.1s\tremaining: 56.5s\n",
      "253:\tlearn: 0.0010826\ttotal: 19.2s\tremaining: 56.4s\n",
      "254:\tlearn: 0.0010794\ttotal: 19.3s\tremaining: 56.3s\n",
      "255:\tlearn: 0.0010770\ttotal: 19.3s\tremaining: 56.2s\n",
      "256:\tlearn: 0.0010754\ttotal: 19.4s\tremaining: 56.1s\n",
      "257:\tlearn: 0.0010731\ttotal: 19.5s\tremaining: 56s\n",
      "258:\tlearn: 0.0010716\ttotal: 19.5s\tremaining: 55.9s\n",
      "259:\tlearn: 0.0010705\ttotal: 19.6s\tremaining: 55.8s\n",
      "260:\tlearn: 0.0010700\ttotal: 19.7s\tremaining: 55.7s\n",
      "261:\tlearn: 0.0010678\ttotal: 19.7s\tremaining: 55.6s\n",
      "262:\tlearn: 0.0010665\ttotal: 19.8s\tremaining: 55.5s\n",
      "263:\tlearn: 0.0010649\ttotal: 19.9s\tremaining: 55.4s\n",
      "264:\tlearn: 0.0010628\ttotal: 19.9s\tremaining: 55.3s\n",
      "265:\tlearn: 0.0010579\ttotal: 20s\tremaining: 55.2s\n",
      "266:\tlearn: 0.0010548\ttotal: 20.1s\tremaining: 55.1s\n",
      "267:\tlearn: 0.0010520\ttotal: 20.2s\tremaining: 55.1s\n",
      "268:\tlearn: 0.0010481\ttotal: 20.2s\tremaining: 55s\n",
      "269:\tlearn: 0.0010471\ttotal: 20.3s\tremaining: 54.9s\n",
      "270:\tlearn: 0.0010459\ttotal: 20.4s\tremaining: 54.8s\n",
      "271:\tlearn: 0.0010439\ttotal: 20.4s\tremaining: 54.7s\n",
      "272:\tlearn: 0.0010427\ttotal: 20.5s\tremaining: 54.6s\n",
      "273:\tlearn: 0.0010423\ttotal: 20.6s\tremaining: 54.5s\n",
      "274:\tlearn: 0.0010409\ttotal: 20.6s\tremaining: 54.4s\n",
      "275:\tlearn: 0.0010397\ttotal: 20.7s\tremaining: 54.3s\n",
      "276:\tlearn: 0.0010361\ttotal: 20.8s\tremaining: 54.2s\n",
      "277:\tlearn: 0.0010337\ttotal: 20.9s\tremaining: 54.2s\n",
      "278:\tlearn: 0.0010322\ttotal: 20.9s\tremaining: 54.1s\n",
      "279:\tlearn: 0.0010274\ttotal: 21s\tremaining: 54s\n",
      "280:\tlearn: 0.0010231\ttotal: 21.1s\tremaining: 53.9s\n",
      "281:\tlearn: 0.0010221\ttotal: 21.1s\tremaining: 53.8s\n",
      "282:\tlearn: 0.0010206\ttotal: 21.2s\tremaining: 53.7s\n",
      "283:\tlearn: 0.0010193\ttotal: 21.3s\tremaining: 53.6s\n",
      "284:\tlearn: 0.0010138\ttotal: 21.3s\tremaining: 53.5s\n",
      "285:\tlearn: 0.0010127\ttotal: 21.4s\tremaining: 53.4s\n",
      "286:\tlearn: 0.0010085\ttotal: 21.5s\tremaining: 53.4s\n",
      "287:\tlearn: 0.0010044\ttotal: 21.6s\tremaining: 53.3s\n",
      "288:\tlearn: 0.0010032\ttotal: 21.6s\tremaining: 53.2s\n",
      "289:\tlearn: 0.0010013\ttotal: 21.7s\tremaining: 53.1s\n",
      "290:\tlearn: 0.0009991\ttotal: 21.8s\tremaining: 53.1s\n",
      "291:\tlearn: 0.0009971\ttotal: 21.8s\tremaining: 53s\n",
      "292:\tlearn: 0.0009961\ttotal: 21.9s\tremaining: 52.9s\n",
      "293:\tlearn: 0.0009859\ttotal: 22s\tremaining: 52.8s\n",
      "294:\tlearn: 0.0009840\ttotal: 22.1s\tremaining: 52.7s\n",
      "295:\tlearn: 0.0009802\ttotal: 22.1s\tremaining: 52.6s\n",
      "296:\tlearn: 0.0009796\ttotal: 22.2s\tremaining: 52.5s\n",
      "297:\tlearn: 0.0009773\ttotal: 22.3s\tremaining: 52.5s\n",
      "298:\tlearn: 0.0009755\ttotal: 22.3s\tremaining: 52.4s\n",
      "299:\tlearn: 0.0009737\ttotal: 22.4s\tremaining: 52.3s\n",
      "300:\tlearn: 0.0009725\ttotal: 22.5s\tremaining: 52.2s\n",
      "301:\tlearn: 0.0009703\ttotal: 22.5s\tremaining: 52.1s\n",
      "302:\tlearn: 0.0009690\ttotal: 22.6s\tremaining: 52s\n",
      "303:\tlearn: 0.0009679\ttotal: 22.7s\tremaining: 51.9s\n",
      "304:\tlearn: 0.0009644\ttotal: 22.8s\tremaining: 51.9s\n",
      "305:\tlearn: 0.0009629\ttotal: 22.8s\tremaining: 51.8s\n",
      "306:\tlearn: 0.0009619\ttotal: 22.9s\tremaining: 51.7s\n",
      "307:\tlearn: 0.0009602\ttotal: 23s\tremaining: 51.6s\n",
      "308:\tlearn: 0.0009585\ttotal: 23s\tremaining: 51.5s\n",
      "309:\tlearn: 0.0009544\ttotal: 23.1s\tremaining: 51.4s\n",
      "310:\tlearn: 0.0009529\ttotal: 23.2s\tremaining: 51.3s\n",
      "311:\tlearn: 0.0009514\ttotal: 23.2s\tremaining: 51.2s\n",
      "312:\tlearn: 0.0009505\ttotal: 23.3s\tremaining: 51.1s\n",
      "313:\tlearn: 0.0009462\ttotal: 23.4s\tremaining: 51.2s\n",
      "314:\tlearn: 0.0009446\ttotal: 23.6s\tremaining: 51.3s\n",
      "315:\tlearn: 0.0009396\ttotal: 23.7s\tremaining: 51.3s\n",
      "316:\tlearn: 0.0009383\ttotal: 23.8s\tremaining: 51.4s\n",
      "317:\tlearn: 0.0009364\ttotal: 24s\tremaining: 51.4s\n",
      "318:\tlearn: 0.0009352\ttotal: 24.1s\tremaining: 51.4s\n",
      "319:\tlearn: 0.0009315\ttotal: 24.1s\tremaining: 51.3s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "320:\tlearn: 0.0009308\ttotal: 24.2s\tremaining: 51.2s\n",
      "321:\tlearn: 0.0009289\ttotal: 24.3s\tremaining: 51.1s\n",
      "322:\tlearn: 0.0009279\ttotal: 24.3s\tremaining: 51s\n",
      "323:\tlearn: 0.0009250\ttotal: 24.4s\tremaining: 50.9s\n",
      "324:\tlearn: 0.0009239\ttotal: 24.5s\tremaining: 50.9s\n",
      "325:\tlearn: 0.0009231\ttotal: 24.6s\tremaining: 50.8s\n",
      "326:\tlearn: 0.0009199\ttotal: 24.6s\tremaining: 50.7s\n",
      "327:\tlearn: 0.0009189\ttotal: 24.7s\tremaining: 50.6s\n",
      "328:\tlearn: 0.0009169\ttotal: 24.8s\tremaining: 50.5s\n",
      "329:\tlearn: 0.0009142\ttotal: 24.8s\tremaining: 50.4s\n",
      "330:\tlearn: 0.0009133\ttotal: 24.9s\tremaining: 50.3s\n",
      "331:\tlearn: 0.0009112\ttotal: 25s\tremaining: 50.2s\n",
      "332:\tlearn: 0.0009101\ttotal: 25s\tremaining: 50.1s\n",
      "333:\tlearn: 0.0009076\ttotal: 25.1s\tremaining: 50s\n",
      "334:\tlearn: 0.0009068\ttotal: 25.2s\tremaining: 50s\n",
      "335:\tlearn: 0.0009060\ttotal: 25.2s\tremaining: 49.9s\n",
      "336:\tlearn: 0.0009027\ttotal: 25.3s\tremaining: 49.8s\n",
      "337:\tlearn: 0.0009007\ttotal: 25.4s\tremaining: 49.7s\n",
      "338:\tlearn: 0.0008999\ttotal: 25.5s\tremaining: 49.6s\n",
      "339:\tlearn: 0.0008990\ttotal: 25.5s\tremaining: 49.6s\n",
      "340:\tlearn: 0.0008974\ttotal: 25.6s\tremaining: 49.5s\n",
      "341:\tlearn: 0.0008955\ttotal: 25.7s\tremaining: 49.4s\n",
      "342:\tlearn: 0.0008929\ttotal: 25.7s\tremaining: 49.3s\n",
      "343:\tlearn: 0.0008897\ttotal: 25.8s\tremaining: 49.2s\n",
      "344:\tlearn: 0.0008809\ttotal: 25.9s\tremaining: 49.1s\n",
      "345:\tlearn: 0.0008783\ttotal: 26s\tremaining: 49.1s\n",
      "346:\tlearn: 0.0008752\ttotal: 26s\tremaining: 49s\n",
      "347:\tlearn: 0.0008735\ttotal: 26.1s\tremaining: 48.9s\n",
      "348:\tlearn: 0.0008722\ttotal: 26.1s\tremaining: 48.8s\n",
      "349:\tlearn: 0.0008701\ttotal: 26.2s\tremaining: 48.7s\n",
      "350:\tlearn: 0.0008682\ttotal: 26.3s\tremaining: 48.6s\n",
      "351:\tlearn: 0.0008668\ttotal: 26.4s\tremaining: 48.5s\n",
      "352:\tlearn: 0.0008644\ttotal: 26.4s\tremaining: 48.4s\n",
      "353:\tlearn: 0.0008630\ttotal: 26.5s\tremaining: 48.4s\n",
      "354:\tlearn: 0.0008597\ttotal: 26.6s\tremaining: 48.3s\n",
      "355:\tlearn: 0.0008594\ttotal: 26.6s\tremaining: 48.2s\n",
      "356:\tlearn: 0.0008581\ttotal: 26.7s\tremaining: 48.1s\n",
      "357:\tlearn: 0.0008576\ttotal: 26.8s\tremaining: 48s\n",
      "358:\tlearn: 0.0008498\ttotal: 26.9s\tremaining: 47.9s\n",
      "359:\tlearn: 0.0008475\ttotal: 26.9s\tremaining: 47.9s\n",
      "360:\tlearn: 0.0008455\ttotal: 27s\tremaining: 47.8s\n",
      "361:\tlearn: 0.0008437\ttotal: 27.1s\tremaining: 47.7s\n",
      "362:\tlearn: 0.0008427\ttotal: 27.1s\tremaining: 47.6s\n",
      "363:\tlearn: 0.0008421\ttotal: 27.2s\tremaining: 47.5s\n",
      "364:\tlearn: 0.0008414\ttotal: 27.3s\tremaining: 47.4s\n",
      "365:\tlearn: 0.0008395\ttotal: 27.3s\tremaining: 47.3s\n",
      "366:\tlearn: 0.0008378\ttotal: 27.4s\tremaining: 47.2s\n",
      "367:\tlearn: 0.0008346\ttotal: 27.5s\tremaining: 47.2s\n",
      "368:\tlearn: 0.0008339\ttotal: 27.5s\tremaining: 47.1s\n",
      "369:\tlearn: 0.0008324\ttotal: 27.6s\tremaining: 47s\n",
      "370:\tlearn: 0.0008308\ttotal: 27.7s\tremaining: 46.9s\n",
      "371:\tlearn: 0.0008301\ttotal: 27.7s\tremaining: 46.8s\n",
      "372:\tlearn: 0.0008285\ttotal: 27.8s\tremaining: 46.7s\n",
      "373:\tlearn: 0.0008270\ttotal: 27.9s\tremaining: 46.7s\n",
      "374:\tlearn: 0.0008259\ttotal: 27.9s\tremaining: 46.6s\n",
      "375:\tlearn: 0.0008229\ttotal: 28s\tremaining: 46.5s\n",
      "376:\tlearn: 0.0008219\ttotal: 28.1s\tremaining: 46.4s\n",
      "377:\tlearn: 0.0008199\ttotal: 28.2s\tremaining: 46.3s\n",
      "378:\tlearn: 0.0008180\ttotal: 28.2s\tremaining: 46.2s\n",
      "379:\tlearn: 0.0008144\ttotal: 28.3s\tremaining: 46.2s\n",
      "380:\tlearn: 0.0008130\ttotal: 28.4s\tremaining: 46.1s\n",
      "381:\tlearn: 0.0008113\ttotal: 28.4s\tremaining: 46s\n",
      "382:\tlearn: 0.0008106\ttotal: 28.5s\tremaining: 45.9s\n",
      "383:\tlearn: 0.0008088\ttotal: 28.6s\tremaining: 45.8s\n",
      "384:\tlearn: 0.0008067\ttotal: 28.6s\tremaining: 45.7s\n",
      "385:\tlearn: 0.0008057\ttotal: 28.7s\tremaining: 45.7s\n",
      "386:\tlearn: 0.0008033\ttotal: 28.8s\tremaining: 45.6s\n",
      "387:\tlearn: 0.0008003\ttotal: 28.8s\tremaining: 45.5s\n",
      "388:\tlearn: 0.0007986\ttotal: 28.9s\tremaining: 45.4s\n",
      "389:\tlearn: 0.0007951\ttotal: 29s\tremaining: 45.3s\n",
      "390:\tlearn: 0.0007943\ttotal: 29.1s\tremaining: 45.3s\n",
      "391:\tlearn: 0.0007923\ttotal: 29.1s\tremaining: 45.2s\n",
      "392:\tlearn: 0.0007908\ttotal: 29.2s\tremaining: 45.1s\n",
      "393:\tlearn: 0.0007865\ttotal: 29.3s\tremaining: 45s\n",
      "394:\tlearn: 0.0007824\ttotal: 29.3s\tremaining: 44.9s\n",
      "395:\tlearn: 0.0007796\ttotal: 29.4s\tremaining: 44.8s\n",
      "396:\tlearn: 0.0007760\ttotal: 29.5s\tremaining: 44.8s\n",
      "397:\tlearn: 0.0007746\ttotal: 29.5s\tremaining: 44.7s\n",
      "398:\tlearn: 0.0007732\ttotal: 29.6s\tremaining: 44.6s\n",
      "399:\tlearn: 0.0007694\ttotal: 29.7s\tremaining: 44.5s\n",
      "400:\tlearn: 0.0007677\ttotal: 29.7s\tremaining: 44.4s\n",
      "401:\tlearn: 0.0007670\ttotal: 29.8s\tremaining: 44.4s\n",
      "402:\tlearn: 0.0007664\ttotal: 29.9s\tremaining: 44.3s\n",
      "403:\tlearn: 0.0007651\ttotal: 29.9s\tremaining: 44.2s\n",
      "404:\tlearn: 0.0007616\ttotal: 30s\tremaining: 44.1s\n",
      "405:\tlearn: 0.0007612\ttotal: 30.1s\tremaining: 44s\n",
      "406:\tlearn: 0.0007607\ttotal: 30.2s\tremaining: 43.9s\n",
      "407:\tlearn: 0.0007583\ttotal: 30.2s\tremaining: 43.9s\n",
      "408:\tlearn: 0.0007575\ttotal: 30.3s\tremaining: 43.8s\n",
      "409:\tlearn: 0.0007556\ttotal: 30.4s\tremaining: 43.7s\n",
      "410:\tlearn: 0.0007549\ttotal: 30.4s\tremaining: 43.6s\n",
      "411:\tlearn: 0.0007485\ttotal: 30.5s\tremaining: 43.5s\n",
      "412:\tlearn: 0.0007465\ttotal: 30.6s\tremaining: 43.4s\n",
      "413:\tlearn: 0.0007442\ttotal: 30.6s\tremaining: 43.4s\n",
      "414:\tlearn: 0.0007382\ttotal: 30.7s\tremaining: 43.3s\n",
      "415:\tlearn: 0.0007329\ttotal: 30.8s\tremaining: 43.2s\n",
      "416:\tlearn: 0.0007304\ttotal: 30.8s\tremaining: 43.1s\n",
      "417:\tlearn: 0.0007296\ttotal: 30.9s\tremaining: 43s\n",
      "418:\tlearn: 0.0007265\ttotal: 31s\tremaining: 43s\n",
      "419:\tlearn: 0.0007247\ttotal: 31s\tremaining: 42.9s\n",
      "420:\tlearn: 0.0007233\ttotal: 31.1s\tremaining: 42.8s\n",
      "421:\tlearn: 0.0007225\ttotal: 31.2s\tremaining: 42.7s\n",
      "422:\tlearn: 0.0007216\ttotal: 31.2s\tremaining: 42.6s\n",
      "423:\tlearn: 0.0007198\ttotal: 31.3s\tremaining: 42.5s\n",
      "424:\tlearn: 0.0007172\ttotal: 31.4s\tremaining: 42.5s\n",
      "425:\tlearn: 0.0007164\ttotal: 31.5s\tremaining: 42.4s\n",
      "426:\tlearn: 0.0007157\ttotal: 31.5s\tremaining: 42.3s\n",
      "427:\tlearn: 0.0007152\ttotal: 31.6s\tremaining: 42.2s\n",
      "428:\tlearn: 0.0007137\ttotal: 31.6s\tremaining: 42.1s\n",
      "429:\tlearn: 0.0007132\ttotal: 31.7s\tremaining: 42s\n",
      "430:\tlearn: 0.0007116\ttotal: 31.8s\tremaining: 42s\n",
      "431:\tlearn: 0.0007110\ttotal: 31.8s\tremaining: 41.9s\n",
      "432:\tlearn: 0.0007093\ttotal: 31.9s\tremaining: 41.8s\n",
      "433:\tlearn: 0.0007059\ttotal: 32s\tremaining: 41.7s\n",
      "434:\tlearn: 0.0007027\ttotal: 32.1s\tremaining: 41.6s\n",
      "435:\tlearn: 0.0007013\ttotal: 32.1s\tremaining: 41.6s\n",
      "436:\tlearn: 0.0007006\ttotal: 32.2s\tremaining: 41.5s\n",
      "437:\tlearn: 0.0006999\ttotal: 32.3s\tremaining: 41.4s\n",
      "438:\tlearn: 0.0006978\ttotal: 32.3s\tremaining: 41.3s\n",
      "439:\tlearn: 0.0006967\ttotal: 32.4s\tremaining: 41.2s\n",
      "440:\tlearn: 0.0006956\ttotal: 32.5s\tremaining: 41.1s\n",
      "441:\tlearn: 0.0006945\ttotal: 32.5s\tremaining: 41.1s\n",
      "442:\tlearn: 0.0006923\ttotal: 32.6s\tremaining: 41s\n",
      "443:\tlearn: 0.0006872\ttotal: 32.7s\tremaining: 40.9s\n",
      "444:\tlearn: 0.0006866\ttotal: 32.7s\tremaining: 40.8s\n",
      "445:\tlearn: 0.0006859\ttotal: 32.8s\tremaining: 40.7s\n",
      "446:\tlearn: 0.0006803\ttotal: 32.9s\tremaining: 40.7s\n",
      "447:\tlearn: 0.0006758\ttotal: 32.9s\tremaining: 40.6s\n",
      "448:\tlearn: 0.0006754\ttotal: 33s\tremaining: 40.5s\n",
      "449:\tlearn: 0.0006733\ttotal: 33.1s\tremaining: 40.4s\n",
      "450:\tlearn: 0.0006672\ttotal: 33.1s\tremaining: 40.3s\n",
      "451:\tlearn: 0.0006648\ttotal: 33.2s\tremaining: 40.3s\n",
      "452:\tlearn: 0.0006628\ttotal: 33.3s\tremaining: 40.2s\n",
      "453:\tlearn: 0.0006606\ttotal: 33.3s\tremaining: 40.1s\n",
      "454:\tlearn: 0.0006565\ttotal: 33.4s\tremaining: 40s\n",
      "455:\tlearn: 0.0006554\ttotal: 33.5s\tremaining: 39.9s\n",
      "456:\tlearn: 0.0006545\ttotal: 33.5s\tremaining: 39.9s\n",
      "457:\tlearn: 0.0006505\ttotal: 33.6s\tremaining: 39.8s\n",
      "458:\tlearn: 0.0006484\ttotal: 33.7s\tremaining: 39.7s\n",
      "459:\tlearn: 0.0006476\ttotal: 33.8s\tremaining: 39.6s\n",
      "460:\tlearn: 0.0006467\ttotal: 33.8s\tremaining: 39.5s\n",
      "461:\tlearn: 0.0006447\ttotal: 33.9s\tremaining: 39.5s\n",
      "462:\tlearn: 0.0006409\ttotal: 34s\tremaining: 39.4s\n",
      "463:\tlearn: 0.0006406\ttotal: 34s\tremaining: 39.3s\n",
      "464:\tlearn: 0.0006384\ttotal: 34.1s\tremaining: 39.2s\n",
      "465:\tlearn: 0.0006351\ttotal: 34.2s\tremaining: 39.2s\n",
      "466:\tlearn: 0.0006342\ttotal: 34.2s\tremaining: 39.1s\n",
      "467:\tlearn: 0.0006293\ttotal: 34.3s\tremaining: 39s\n",
      "468:\tlearn: 0.0006290\ttotal: 34.5s\tremaining: 39.1s\n",
      "469:\tlearn: 0.0006281\ttotal: 34.6s\tremaining: 39.1s\n",
      "470:\tlearn: 0.0006274\ttotal: 34.8s\tremaining: 39s\n",
      "471:\tlearn: 0.0006250\ttotal: 34.9s\tremaining: 39s\n",
      "472:\tlearn: 0.0006241\ttotal: 35s\tremaining: 39s\n",
      "473:\tlearn: 0.0006232\ttotal: 35.1s\tremaining: 39s\n",
      "474:\tlearn: 0.0006213\ttotal: 35.2s\tremaining: 38.9s\n",
      "475:\tlearn: 0.0006205\ttotal: 35.2s\tremaining: 38.8s\n",
      "476:\tlearn: 0.0006199\ttotal: 35.3s\tremaining: 38.7s\n",
      "477:\tlearn: 0.0006190\ttotal: 35.4s\tremaining: 38.7s\n",
      "478:\tlearn: 0.0006182\ttotal: 35.5s\tremaining: 38.6s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "479:\tlearn: 0.0006178\ttotal: 35.5s\tremaining: 38.5s\n",
      "480:\tlearn: 0.0006172\ttotal: 35.6s\tremaining: 38.4s\n",
      "481:\tlearn: 0.0006160\ttotal: 35.7s\tremaining: 38.3s\n",
      "482:\tlearn: 0.0006153\ttotal: 35.7s\tremaining: 38.3s\n",
      "483:\tlearn: 0.0006145\ttotal: 35.8s\tremaining: 38.2s\n",
      "484:\tlearn: 0.0006097\ttotal: 35.9s\tremaining: 38.1s\n",
      "485:\tlearn: 0.0006084\ttotal: 35.9s\tremaining: 38s\n",
      "486:\tlearn: 0.0006076\ttotal: 36s\tremaining: 37.9s\n",
      "487:\tlearn: 0.0006055\ttotal: 36.1s\tremaining: 37.9s\n",
      "488:\tlearn: 0.0006049\ttotal: 36.1s\tremaining: 37.8s\n",
      "489:\tlearn: 0.0006020\ttotal: 36.2s\tremaining: 37.7s\n",
      "490:\tlearn: 0.0006013\ttotal: 36.3s\tremaining: 37.6s\n",
      "491:\tlearn: 0.0005960\ttotal: 36.4s\tremaining: 37.5s\n",
      "492:\tlearn: 0.0005948\ttotal: 36.4s\tremaining: 37.5s\n",
      "493:\tlearn: 0.0005929\ttotal: 36.5s\tremaining: 37.4s\n",
      "494:\tlearn: 0.0005926\ttotal: 36.5s\tremaining: 37.3s\n",
      "495:\tlearn: 0.0005922\ttotal: 36.6s\tremaining: 37.2s\n",
      "496:\tlearn: 0.0005894\ttotal: 36.7s\tremaining: 37.1s\n",
      "497:\tlearn: 0.0005888\ttotal: 36.8s\tremaining: 37s\n",
      "498:\tlearn: 0.0005878\ttotal: 36.8s\tremaining: 37s\n",
      "499:\tlearn: 0.0005870\ttotal: 36.9s\tremaining: 36.9s\n",
      "500:\tlearn: 0.0005862\ttotal: 37s\tremaining: 36.8s\n",
      "501:\tlearn: 0.0005856\ttotal: 37s\tremaining: 36.7s\n",
      "502:\tlearn: 0.0005826\ttotal: 37.1s\tremaining: 36.7s\n",
      "503:\tlearn: 0.0005803\ttotal: 37.2s\tremaining: 36.6s\n",
      "504:\tlearn: 0.0005790\ttotal: 37.2s\tremaining: 36.5s\n",
      "505:\tlearn: 0.0005769\ttotal: 37.3s\tremaining: 36.4s\n",
      "506:\tlearn: 0.0005763\ttotal: 37.4s\tremaining: 36.3s\n",
      "507:\tlearn: 0.0005750\ttotal: 37.4s\tremaining: 36.3s\n",
      "508:\tlearn: 0.0005738\ttotal: 37.5s\tremaining: 36.2s\n",
      "509:\tlearn: 0.0005729\ttotal: 37.6s\tremaining: 36.1s\n",
      "510:\tlearn: 0.0005726\ttotal: 37.6s\tremaining: 36s\n",
      "511:\tlearn: 0.0005713\ttotal: 37.7s\tremaining: 35.9s\n",
      "512:\tlearn: 0.0005701\ttotal: 37.8s\tremaining: 35.9s\n",
      "513:\tlearn: 0.0005690\ttotal: 37.8s\tremaining: 35.8s\n",
      "514:\tlearn: 0.0005688\ttotal: 37.9s\tremaining: 35.7s\n",
      "515:\tlearn: 0.0005677\ttotal: 38s\tremaining: 35.6s\n",
      "516:\tlearn: 0.0005664\ttotal: 38s\tremaining: 35.5s\n",
      "517:\tlearn: 0.0005650\ttotal: 38.1s\tremaining: 35.5s\n",
      "518:\tlearn: 0.0005633\ttotal: 38.2s\tremaining: 35.4s\n",
      "519:\tlearn: 0.0005623\ttotal: 38.3s\tremaining: 35.3s\n",
      "520:\tlearn: 0.0005614\ttotal: 38.3s\tremaining: 35.2s\n",
      "521:\tlearn: 0.0005609\ttotal: 38.4s\tremaining: 35.2s\n",
      "522:\tlearn: 0.0005607\ttotal: 38.5s\tremaining: 35.1s\n",
      "523:\tlearn: 0.0005598\ttotal: 38.5s\tremaining: 35s\n",
      "524:\tlearn: 0.0005592\ttotal: 38.6s\tremaining: 34.9s\n",
      "525:\tlearn: 0.0005587\ttotal: 38.7s\tremaining: 34.8s\n",
      "526:\tlearn: 0.0005551\ttotal: 38.7s\tremaining: 34.8s\n",
      "527:\tlearn: 0.0005539\ttotal: 38.8s\tremaining: 34.7s\n",
      "528:\tlearn: 0.0005528\ttotal: 38.9s\tremaining: 34.6s\n",
      "529:\tlearn: 0.0005517\ttotal: 38.9s\tremaining: 34.5s\n",
      "530:\tlearn: 0.0005502\ttotal: 39s\tremaining: 34.5s\n",
      "531:\tlearn: 0.0005495\ttotal: 39.1s\tremaining: 34.4s\n",
      "532:\tlearn: 0.0005490\ttotal: 39.1s\tremaining: 34.3s\n",
      "533:\tlearn: 0.0005471\ttotal: 39.2s\tremaining: 34.2s\n",
      "534:\tlearn: 0.0005456\ttotal: 39.3s\tremaining: 34.1s\n",
      "535:\tlearn: 0.0005442\ttotal: 39.4s\tremaining: 34.1s\n",
      "536:\tlearn: 0.0005433\ttotal: 39.4s\tremaining: 34s\n",
      "537:\tlearn: 0.0005429\ttotal: 39.5s\tremaining: 33.9s\n",
      "538:\tlearn: 0.0005410\ttotal: 39.6s\tremaining: 33.8s\n",
      "539:\tlearn: 0.0005394\ttotal: 39.6s\tremaining: 33.8s\n",
      "540:\tlearn: 0.0005385\ttotal: 39.7s\tremaining: 33.7s\n",
      "541:\tlearn: 0.0005377\ttotal: 39.8s\tremaining: 33.6s\n",
      "542:\tlearn: 0.0005367\ttotal: 39.8s\tremaining: 33.5s\n",
      "543:\tlearn: 0.0005356\ttotal: 39.9s\tremaining: 33.5s\n",
      "544:\tlearn: 0.0005333\ttotal: 40s\tremaining: 33.4s\n",
      "545:\tlearn: 0.0005323\ttotal: 40.1s\tremaining: 33.3s\n",
      "546:\tlearn: 0.0005307\ttotal: 40.1s\tremaining: 33.2s\n",
      "547:\tlearn: 0.0005272\ttotal: 40.2s\tremaining: 33.2s\n",
      "548:\tlearn: 0.0005257\ttotal: 40.3s\tremaining: 33.1s\n",
      "549:\tlearn: 0.0005243\ttotal: 40.4s\tremaining: 33s\n",
      "550:\tlearn: 0.0005228\ttotal: 40.4s\tremaining: 32.9s\n",
      "551:\tlearn: 0.0005221\ttotal: 40.5s\tremaining: 32.9s\n",
      "552:\tlearn: 0.0005190\ttotal: 40.6s\tremaining: 32.8s\n",
      "553:\tlearn: 0.0005187\ttotal: 40.6s\tremaining: 32.7s\n",
      "554:\tlearn: 0.0005160\ttotal: 40.7s\tremaining: 32.6s\n",
      "555:\tlearn: 0.0005151\ttotal: 40.8s\tremaining: 32.6s\n",
      "556:\tlearn: 0.0005128\ttotal: 40.9s\tremaining: 32.5s\n",
      "557:\tlearn: 0.0005099\ttotal: 40.9s\tremaining: 32.4s\n",
      "558:\tlearn: 0.0005066\ttotal: 41s\tremaining: 32.3s\n",
      "559:\tlearn: 0.0005062\ttotal: 41.1s\tremaining: 32.3s\n",
      "560:\tlearn: 0.0005052\ttotal: 41.1s\tremaining: 32.2s\n",
      "561:\tlearn: 0.0005045\ttotal: 41.2s\tremaining: 32.1s\n",
      "562:\tlearn: 0.0005027\ttotal: 41.3s\tremaining: 32s\n",
      "563:\tlearn: 0.0005020\ttotal: 41.3s\tremaining: 32s\n",
      "564:\tlearn: 0.0005013\ttotal: 41.4s\tremaining: 31.9s\n",
      "565:\tlearn: 0.0004994\ttotal: 41.5s\tremaining: 31.8s\n",
      "566:\tlearn: 0.0004973\ttotal: 41.6s\tremaining: 31.7s\n",
      "567:\tlearn: 0.0004969\ttotal: 41.6s\tremaining: 31.7s\n",
      "568:\tlearn: 0.0004965\ttotal: 41.7s\tremaining: 31.6s\n",
      "569:\tlearn: 0.0004959\ttotal: 41.8s\tremaining: 31.5s\n",
      "570:\tlearn: 0.0004952\ttotal: 41.8s\tremaining: 31.4s\n",
      "571:\tlearn: 0.0004939\ttotal: 41.9s\tremaining: 31.4s\n",
      "572:\tlearn: 0.0004930\ttotal: 42s\tremaining: 31.3s\n",
      "573:\tlearn: 0.0004905\ttotal: 42s\tremaining: 31.2s\n",
      "574:\tlearn: 0.0004898\ttotal: 42.1s\tremaining: 31.1s\n",
      "575:\tlearn: 0.0004885\ttotal: 42.2s\tremaining: 31.1s\n",
      "576:\tlearn: 0.0004874\ttotal: 42.3s\tremaining: 31s\n",
      "577:\tlearn: 0.0004868\ttotal: 42.3s\tremaining: 30.9s\n",
      "578:\tlearn: 0.0004856\ttotal: 42.4s\tremaining: 30.8s\n",
      "579:\tlearn: 0.0004835\ttotal: 42.5s\tremaining: 30.8s\n",
      "580:\tlearn: 0.0004825\ttotal: 42.5s\tremaining: 30.7s\n",
      "581:\tlearn: 0.0004819\ttotal: 42.6s\tremaining: 30.6s\n",
      "582:\tlearn: 0.0004800\ttotal: 42.7s\tremaining: 30.5s\n",
      "583:\tlearn: 0.0004790\ttotal: 42.7s\tremaining: 30.4s\n",
      "584:\tlearn: 0.0004779\ttotal: 42.8s\tremaining: 30.4s\n",
      "585:\tlearn: 0.0004776\ttotal: 42.9s\tremaining: 30.3s\n",
      "586:\tlearn: 0.0004749\ttotal: 43s\tremaining: 30.2s\n",
      "587:\tlearn: 0.0004745\ttotal: 43s\tremaining: 30.2s\n",
      "588:\tlearn: 0.0004738\ttotal: 43.1s\tremaining: 30.1s\n",
      "589:\tlearn: 0.0004722\ttotal: 43.2s\tremaining: 30s\n",
      "590:\tlearn: 0.0004710\ttotal: 43.2s\tremaining: 29.9s\n",
      "591:\tlearn: 0.0004707\ttotal: 43.3s\tremaining: 29.9s\n",
      "592:\tlearn: 0.0004699\ttotal: 43.5s\tremaining: 29.8s\n",
      "593:\tlearn: 0.0004683\ttotal: 43.6s\tremaining: 29.8s\n",
      "594:\tlearn: 0.0004674\ttotal: 43.7s\tremaining: 29.8s\n",
      "595:\tlearn: 0.0004670\ttotal: 43.9s\tremaining: 29.7s\n",
      "596:\tlearn: 0.0004664\ttotal: 44s\tremaining: 29.7s\n",
      "597:\tlearn: 0.0004655\ttotal: 44.1s\tremaining: 29.7s\n",
      "598:\tlearn: 0.0004652\ttotal: 44.3s\tremaining: 29.6s\n",
      "599:\tlearn: 0.0004648\ttotal: 44.4s\tremaining: 29.6s\n",
      "600:\tlearn: 0.0004646\ttotal: 44.5s\tremaining: 29.6s\n",
      "601:\tlearn: 0.0004643\ttotal: 44.6s\tremaining: 29.5s\n",
      "602:\tlearn: 0.0004623\ttotal: 44.7s\tremaining: 29.4s\n",
      "603:\tlearn: 0.0004603\ttotal: 44.7s\tremaining: 29.3s\n",
      "604:\tlearn: 0.0004581\ttotal: 44.8s\tremaining: 29.3s\n",
      "605:\tlearn: 0.0004575\ttotal: 44.9s\tremaining: 29.2s\n",
      "606:\tlearn: 0.0004570\ttotal: 45s\tremaining: 29.1s\n",
      "607:\tlearn: 0.0004566\ttotal: 45s\tremaining: 29s\n",
      "608:\tlearn: 0.0004560\ttotal: 45.1s\tremaining: 29s\n",
      "609:\tlearn: 0.0004555\ttotal: 45.2s\tremaining: 28.9s\n",
      "610:\tlearn: 0.0004549\ttotal: 45.2s\tremaining: 28.8s\n",
      "611:\tlearn: 0.0004526\ttotal: 45.4s\tremaining: 28.8s\n",
      "612:\tlearn: 0.0004521\ttotal: 45.6s\tremaining: 28.8s\n",
      "613:\tlearn: 0.0004508\ttotal: 45.7s\tremaining: 28.7s\n",
      "614:\tlearn: 0.0004497\ttotal: 45.8s\tremaining: 28.7s\n",
      "615:\tlearn: 0.0004494\ttotal: 45.9s\tremaining: 28.6s\n",
      "616:\tlearn: 0.0004470\ttotal: 46.1s\tremaining: 28.6s\n",
      "617:\tlearn: 0.0004464\ttotal: 46.1s\tremaining: 28.5s\n",
      "618:\tlearn: 0.0004460\ttotal: 46.2s\tremaining: 28.4s\n",
      "619:\tlearn: 0.0004444\ttotal: 46.3s\tremaining: 28.4s\n",
      "620:\tlearn: 0.0004442\ttotal: 46.3s\tremaining: 28.3s\n",
      "621:\tlearn: 0.0004438\ttotal: 46.4s\tremaining: 28.2s\n",
      "622:\tlearn: 0.0004434\ttotal: 46.5s\tremaining: 28.1s\n",
      "623:\tlearn: 0.0004418\ttotal: 46.5s\tremaining: 28s\n",
      "624:\tlearn: 0.0004413\ttotal: 46.6s\tremaining: 28s\n",
      "625:\tlearn: 0.0004396\ttotal: 46.7s\tremaining: 27.9s\n",
      "626:\tlearn: 0.0004386\ttotal: 46.8s\tremaining: 27.8s\n",
      "627:\tlearn: 0.0004373\ttotal: 46.8s\tremaining: 27.7s\n",
      "628:\tlearn: 0.0004359\ttotal: 46.9s\tremaining: 27.7s\n",
      "629:\tlearn: 0.0004357\ttotal: 47s\tremaining: 27.6s\n",
      "630:\tlearn: 0.0004320\ttotal: 47s\tremaining: 27.5s\n",
      "631:\tlearn: 0.0004312\ttotal: 47.1s\tremaining: 27.4s\n",
      "632:\tlearn: 0.0004308\ttotal: 47.2s\tremaining: 27.4s\n",
      "633:\tlearn: 0.0004306\ttotal: 47.2s\tremaining: 27.3s\n",
      "634:\tlearn: 0.0004298\ttotal: 47.3s\tremaining: 27.2s\n",
      "635:\tlearn: 0.0004287\ttotal: 47.4s\tremaining: 27.1s\n",
      "636:\tlearn: 0.0004283\ttotal: 47.5s\tremaining: 27s\n",
      "637:\tlearn: 0.0004280\ttotal: 47.5s\tremaining: 27s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "638:\tlearn: 0.0004276\ttotal: 47.6s\tremaining: 26.9s\n",
      "639:\tlearn: 0.0004268\ttotal: 47.7s\tremaining: 26.8s\n",
      "640:\tlearn: 0.0004265\ttotal: 47.7s\tremaining: 26.7s\n",
      "641:\tlearn: 0.0004262\ttotal: 47.8s\tremaining: 26.7s\n",
      "642:\tlearn: 0.0004237\ttotal: 47.9s\tremaining: 26.6s\n",
      "643:\tlearn: 0.0004235\ttotal: 47.9s\tremaining: 26.5s\n",
      "644:\tlearn: 0.0004228\ttotal: 48s\tremaining: 26.4s\n",
      "645:\tlearn: 0.0004218\ttotal: 48.1s\tremaining: 26.3s\n",
      "646:\tlearn: 0.0004215\ttotal: 48.1s\tremaining: 26.3s\n",
      "647:\tlearn: 0.0004191\ttotal: 48.2s\tremaining: 26.2s\n",
      "648:\tlearn: 0.0004186\ttotal: 48.3s\tremaining: 26.1s\n",
      "649:\tlearn: 0.0004184\ttotal: 48.4s\tremaining: 26s\n",
      "650:\tlearn: 0.0004179\ttotal: 48.4s\tremaining: 26s\n",
      "651:\tlearn: 0.0004176\ttotal: 48.5s\tremaining: 25.9s\n",
      "652:\tlearn: 0.0004166\ttotal: 48.6s\tremaining: 25.8s\n",
      "653:\tlearn: 0.0004146\ttotal: 48.6s\tremaining: 25.7s\n",
      "654:\tlearn: 0.0004131\ttotal: 48.7s\tremaining: 25.7s\n",
      "655:\tlearn: 0.0004129\ttotal: 48.8s\tremaining: 25.6s\n",
      "656:\tlearn: 0.0004126\ttotal: 48.9s\tremaining: 25.5s\n",
      "657:\tlearn: 0.0004115\ttotal: 48.9s\tremaining: 25.4s\n",
      "658:\tlearn: 0.0004109\ttotal: 49s\tremaining: 25.4s\n",
      "659:\tlearn: 0.0004106\ttotal: 49.1s\tremaining: 25.3s\n",
      "660:\tlearn: 0.0004100\ttotal: 49.1s\tremaining: 25.2s\n",
      "661:\tlearn: 0.0004092\ttotal: 49.2s\tremaining: 25.1s\n",
      "662:\tlearn: 0.0004084\ttotal: 49.3s\tremaining: 25s\n",
      "663:\tlearn: 0.0004082\ttotal: 49.3s\tremaining: 25s\n",
      "664:\tlearn: 0.0004062\ttotal: 49.4s\tremaining: 24.9s\n",
      "665:\tlearn: 0.0004043\ttotal: 49.5s\tremaining: 24.8s\n",
      "666:\tlearn: 0.0004039\ttotal: 49.5s\tremaining: 24.7s\n",
      "667:\tlearn: 0.0004036\ttotal: 49.6s\tremaining: 24.7s\n",
      "668:\tlearn: 0.0004022\ttotal: 49.7s\tremaining: 24.6s\n",
      "669:\tlearn: 0.0004012\ttotal: 49.8s\tremaining: 24.5s\n",
      "670:\tlearn: 0.0003986\ttotal: 49.8s\tremaining: 24.4s\n",
      "671:\tlearn: 0.0003969\ttotal: 49.9s\tremaining: 24.4s\n",
      "672:\tlearn: 0.0003950\ttotal: 50s\tremaining: 24.3s\n",
      "673:\tlearn: 0.0003935\ttotal: 50s\tremaining: 24.2s\n",
      "674:\tlearn: 0.0003932\ttotal: 50.1s\tremaining: 24.1s\n",
      "675:\tlearn: 0.0003908\ttotal: 50.2s\tremaining: 24.1s\n",
      "676:\tlearn: 0.0003906\ttotal: 50.2s\tremaining: 24s\n",
      "677:\tlearn: 0.0003899\ttotal: 50.3s\tremaining: 23.9s\n",
      "678:\tlearn: 0.0003896\ttotal: 50.4s\tremaining: 23.8s\n",
      "679:\tlearn: 0.0003891\ttotal: 50.5s\tremaining: 23.7s\n",
      "680:\tlearn: 0.0003860\ttotal: 50.5s\tremaining: 23.7s\n",
      "681:\tlearn: 0.0003852\ttotal: 50.6s\tremaining: 23.6s\n",
      "682:\tlearn: 0.0003835\ttotal: 50.7s\tremaining: 23.5s\n",
      "683:\tlearn: 0.0003817\ttotal: 50.7s\tremaining: 23.4s\n",
      "684:\tlearn: 0.0003808\ttotal: 50.8s\tremaining: 23.4s\n",
      "685:\tlearn: 0.0003806\ttotal: 50.9s\tremaining: 23.3s\n",
      "686:\tlearn: 0.0003784\ttotal: 51s\tremaining: 23.2s\n",
      "687:\tlearn: 0.0003782\ttotal: 51s\tremaining: 23.1s\n",
      "688:\tlearn: 0.0003777\ttotal: 51.1s\tremaining: 23.1s\n",
      "689:\tlearn: 0.0003750\ttotal: 51.2s\tremaining: 23s\n",
      "690:\tlearn: 0.0003743\ttotal: 51.2s\tremaining: 22.9s\n",
      "691:\tlearn: 0.0003737\ttotal: 51.3s\tremaining: 22.8s\n",
      "692:\tlearn: 0.0003733\ttotal: 51.4s\tremaining: 22.8s\n",
      "693:\tlearn: 0.0003724\ttotal: 51.4s\tremaining: 22.7s\n",
      "694:\tlearn: 0.0003703\ttotal: 51.5s\tremaining: 22.6s\n",
      "695:\tlearn: 0.0003702\ttotal: 51.6s\tremaining: 22.5s\n",
      "696:\tlearn: 0.0003696\ttotal: 51.6s\tremaining: 22.4s\n",
      "697:\tlearn: 0.0003694\ttotal: 51.7s\tremaining: 22.4s\n",
      "698:\tlearn: 0.0003678\ttotal: 51.8s\tremaining: 22.3s\n",
      "699:\tlearn: 0.0003674\ttotal: 51.8s\tremaining: 22.2s\n",
      "700:\tlearn: 0.0003672\ttotal: 51.9s\tremaining: 22.1s\n",
      "701:\tlearn: 0.0003670\ttotal: 52s\tremaining: 22.1s\n",
      "702:\tlearn: 0.0003669\ttotal: 52s\tremaining: 22s\n",
      "703:\tlearn: 0.0003667\ttotal: 52.1s\tremaining: 21.9s\n",
      "704:\tlearn: 0.0003658\ttotal: 52.2s\tremaining: 21.8s\n",
      "705:\tlearn: 0.0003621\ttotal: 52.2s\tremaining: 21.8s\n",
      "706:\tlearn: 0.0003611\ttotal: 52.3s\tremaining: 21.7s\n",
      "707:\tlearn: 0.0003608\ttotal: 52.4s\tremaining: 21.6s\n",
      "708:\tlearn: 0.0003605\ttotal: 52.4s\tremaining: 21.5s\n",
      "709:\tlearn: 0.0003589\ttotal: 52.5s\tremaining: 21.4s\n",
      "710:\tlearn: 0.0003586\ttotal: 52.6s\tremaining: 21.4s\n",
      "711:\tlearn: 0.0003570\ttotal: 52.6s\tremaining: 21.3s\n",
      "712:\tlearn: 0.0003569\ttotal: 52.7s\tremaining: 21.2s\n",
      "713:\tlearn: 0.0003567\ttotal: 52.8s\tremaining: 21.1s\n",
      "714:\tlearn: 0.0003557\ttotal: 52.9s\tremaining: 21.1s\n",
      "715:\tlearn: 0.0003554\ttotal: 52.9s\tremaining: 21s\n",
      "716:\tlearn: 0.0003550\ttotal: 53s\tremaining: 20.9s\n",
      "717:\tlearn: 0.0003535\ttotal: 53.1s\tremaining: 20.8s\n",
      "718:\tlearn: 0.0003531\ttotal: 53.1s\tremaining: 20.8s\n",
      "719:\tlearn: 0.0003521\ttotal: 53.2s\tremaining: 20.7s\n",
      "720:\tlearn: 0.0003520\ttotal: 53.3s\tremaining: 20.6s\n",
      "721:\tlearn: 0.0003513\ttotal: 53.3s\tremaining: 20.5s\n",
      "722:\tlearn: 0.0003506\ttotal: 53.4s\tremaining: 20.5s\n",
      "723:\tlearn: 0.0003493\ttotal: 53.5s\tremaining: 20.4s\n",
      "724:\tlearn: 0.0003491\ttotal: 53.6s\tremaining: 20.3s\n",
      "725:\tlearn: 0.0003484\ttotal: 53.6s\tremaining: 20.2s\n",
      "726:\tlearn: 0.0003475\ttotal: 53.7s\tremaining: 20.2s\n",
      "727:\tlearn: 0.0003471\ttotal: 53.8s\tremaining: 20.1s\n",
      "728:\tlearn: 0.0003451\ttotal: 53.8s\tremaining: 20s\n",
      "729:\tlearn: 0.0003446\ttotal: 53.9s\tremaining: 19.9s\n",
      "730:\tlearn: 0.0003432\ttotal: 54s\tremaining: 19.9s\n",
      "731:\tlearn: 0.0003421\ttotal: 54.1s\tremaining: 19.8s\n",
      "732:\tlearn: 0.0003418\ttotal: 54.1s\tremaining: 19.7s\n",
      "733:\tlearn: 0.0003414\ttotal: 54.2s\tremaining: 19.6s\n",
      "734:\tlearn: 0.0003403\ttotal: 54.3s\tremaining: 19.6s\n",
      "735:\tlearn: 0.0003400\ttotal: 54.3s\tremaining: 19.5s\n",
      "736:\tlearn: 0.0003399\ttotal: 54.4s\tremaining: 19.4s\n",
      "737:\tlearn: 0.0003391\ttotal: 54.5s\tremaining: 19.3s\n",
      "738:\tlearn: 0.0003377\ttotal: 54.5s\tremaining: 19.3s\n",
      "739:\tlearn: 0.0003369\ttotal: 54.6s\tremaining: 19.2s\n",
      "740:\tlearn: 0.0003361\ttotal: 54.7s\tremaining: 19.1s\n",
      "741:\tlearn: 0.0003353\ttotal: 54.7s\tremaining: 19s\n",
      "742:\tlearn: 0.0003348\ttotal: 54.8s\tremaining: 19s\n",
      "743:\tlearn: 0.0003346\ttotal: 54.9s\tremaining: 18.9s\n",
      "744:\tlearn: 0.0003344\ttotal: 54.9s\tremaining: 18.8s\n",
      "745:\tlearn: 0.0003325\ttotal: 55s\tremaining: 18.7s\n",
      "746:\tlearn: 0.0003323\ttotal: 55.1s\tremaining: 18.7s\n",
      "747:\tlearn: 0.0003321\ttotal: 55.1s\tremaining: 18.6s\n",
      "748:\tlearn: 0.0003310\ttotal: 55.2s\tremaining: 18.5s\n",
      "749:\tlearn: 0.0003306\ttotal: 55.3s\tremaining: 18.4s\n",
      "750:\tlearn: 0.0003294\ttotal: 55.4s\tremaining: 18.4s\n",
      "751:\tlearn: 0.0003281\ttotal: 55.4s\tremaining: 18.3s\n",
      "752:\tlearn: 0.0003257\ttotal: 55.5s\tremaining: 18.2s\n",
      "753:\tlearn: 0.0003252\ttotal: 55.6s\tremaining: 18.1s\n",
      "754:\tlearn: 0.0003246\ttotal: 55.7s\tremaining: 18.1s\n",
      "755:\tlearn: 0.0003245\ttotal: 55.7s\tremaining: 18s\n",
      "756:\tlearn: 0.0003228\ttotal: 55.8s\tremaining: 17.9s\n",
      "757:\tlearn: 0.0003218\ttotal: 55.9s\tremaining: 17.8s\n",
      "758:\tlearn: 0.0003214\ttotal: 55.9s\tremaining: 17.8s\n",
      "759:\tlearn: 0.0003186\ttotal: 56s\tremaining: 17.7s\n",
      "760:\tlearn: 0.0003181\ttotal: 56.1s\tremaining: 17.6s\n",
      "761:\tlearn: 0.0003167\ttotal: 56.2s\tremaining: 17.5s\n",
      "762:\tlearn: 0.0003166\ttotal: 56.2s\tremaining: 17.5s\n",
      "763:\tlearn: 0.0003159\ttotal: 56.4s\tremaining: 17.4s\n",
      "764:\tlearn: 0.0003157\ttotal: 56.5s\tremaining: 17.4s\n",
      "765:\tlearn: 0.0003156\ttotal: 56.6s\tremaining: 17.3s\n",
      "766:\tlearn: 0.0003155\ttotal: 56.8s\tremaining: 17.2s\n",
      "767:\tlearn: 0.0003152\ttotal: 56.9s\tremaining: 17.2s\n",
      "768:\tlearn: 0.0003141\ttotal: 57s\tremaining: 17.1s\n",
      "769:\tlearn: 0.0003119\ttotal: 57.1s\tremaining: 17s\n",
      "770:\tlearn: 0.0003117\ttotal: 57.1s\tremaining: 17s\n",
      "771:\tlearn: 0.0003114\ttotal: 57.2s\tremaining: 16.9s\n",
      "772:\tlearn: 0.0003085\ttotal: 57.3s\tremaining: 16.8s\n",
      "773:\tlearn: 0.0003069\ttotal: 57.3s\tremaining: 16.7s\n",
      "774:\tlearn: 0.0003058\ttotal: 57.4s\tremaining: 16.7s\n",
      "775:\tlearn: 0.0003053\ttotal: 57.5s\tremaining: 16.6s\n",
      "776:\tlearn: 0.0003048\ttotal: 57.6s\tremaining: 16.5s\n",
      "777:\tlearn: 0.0003047\ttotal: 57.6s\tremaining: 16.4s\n",
      "778:\tlearn: 0.0003043\ttotal: 57.7s\tremaining: 16.4s\n",
      "779:\tlearn: 0.0003040\ttotal: 57.8s\tremaining: 16.3s\n",
      "780:\tlearn: 0.0003037\ttotal: 57.8s\tremaining: 16.2s\n",
      "781:\tlearn: 0.0003021\ttotal: 57.9s\tremaining: 16.1s\n",
      "782:\tlearn: 0.0003020\ttotal: 58s\tremaining: 16.1s\n",
      "783:\tlearn: 0.0003016\ttotal: 58s\tremaining: 16s\n",
      "784:\tlearn: 0.0003014\ttotal: 58.1s\tremaining: 15.9s\n",
      "785:\tlearn: 0.0003002\ttotal: 58.2s\tremaining: 15.8s\n",
      "786:\tlearn: 0.0003000\ttotal: 58.2s\tremaining: 15.8s\n",
      "787:\tlearn: 0.0002988\ttotal: 58.3s\tremaining: 15.7s\n",
      "788:\tlearn: 0.0002983\ttotal: 58.4s\tremaining: 15.6s\n",
      "789:\tlearn: 0.0002959\ttotal: 58.4s\tremaining: 15.5s\n",
      "790:\tlearn: 0.0002957\ttotal: 58.5s\tremaining: 15.5s\n",
      "791:\tlearn: 0.0002944\ttotal: 58.6s\tremaining: 15.4s\n",
      "792:\tlearn: 0.0002933\ttotal: 58.6s\tremaining: 15.3s\n",
      "793:\tlearn: 0.0002931\ttotal: 58.7s\tremaining: 15.2s\n",
      "794:\tlearn: 0.0002928\ttotal: 58.8s\tremaining: 15.2s\n",
      "795:\tlearn: 0.0002927\ttotal: 58.8s\tremaining: 15.1s\n",
      "796:\tlearn: 0.0002925\ttotal: 58.9s\tremaining: 15s\n",
      "797:\tlearn: 0.0002921\ttotal: 59s\tremaining: 14.9s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "798:\tlearn: 0.0002911\ttotal: 59s\tremaining: 14.9s\n",
      "799:\tlearn: 0.0002910\ttotal: 59.1s\tremaining: 14.8s\n",
      "800:\tlearn: 0.0002902\ttotal: 59.2s\tremaining: 14.7s\n",
      "801:\tlearn: 0.0002900\ttotal: 59.2s\tremaining: 14.6s\n",
      "802:\tlearn: 0.0002893\ttotal: 59.3s\tremaining: 14.6s\n",
      "803:\tlearn: 0.0002887\ttotal: 59.4s\tremaining: 14.5s\n",
      "804:\tlearn: 0.0002883\ttotal: 59.4s\tremaining: 14.4s\n",
      "805:\tlearn: 0.0002867\ttotal: 59.5s\tremaining: 14.3s\n",
      "806:\tlearn: 0.0002866\ttotal: 59.6s\tremaining: 14.2s\n",
      "807:\tlearn: 0.0002865\ttotal: 59.6s\tremaining: 14.2s\n",
      "808:\tlearn: 0.0002855\ttotal: 59.7s\tremaining: 14.1s\n",
      "809:\tlearn: 0.0002837\ttotal: 59.8s\tremaining: 14s\n",
      "810:\tlearn: 0.0002833\ttotal: 59.9s\tremaining: 13.9s\n",
      "811:\tlearn: 0.0002828\ttotal: 59.9s\tremaining: 13.9s\n",
      "812:\tlearn: 0.0002827\ttotal: 60s\tremaining: 13.8s\n",
      "813:\tlearn: 0.0002816\ttotal: 1m\tremaining: 13.7s\n",
      "814:\tlearn: 0.0002815\ttotal: 1m\tremaining: 13.7s\n",
      "815:\tlearn: 0.0002806\ttotal: 1m\tremaining: 13.6s\n",
      "816:\tlearn: 0.0002794\ttotal: 1m\tremaining: 13.5s\n",
      "817:\tlearn: 0.0002792\ttotal: 1m\tremaining: 13.4s\n",
      "818:\tlearn: 0.0002791\ttotal: 1m\tremaining: 13.4s\n",
      "819:\tlearn: 0.0002785\ttotal: 1m\tremaining: 13.3s\n",
      "820:\tlearn: 0.0002766\ttotal: 1m\tremaining: 13.2s\n",
      "821:\tlearn: 0.0002760\ttotal: 1m\tremaining: 13.1s\n",
      "822:\tlearn: 0.0002759\ttotal: 1m\tremaining: 13.1s\n",
      "823:\tlearn: 0.0002751\ttotal: 1m\tremaining: 13s\n",
      "824:\tlearn: 0.0002749\ttotal: 1m\tremaining: 12.9s\n",
      "825:\tlearn: 0.0002744\ttotal: 1m\tremaining: 12.8s\n",
      "826:\tlearn: 0.0002743\ttotal: 1m\tremaining: 12.8s\n",
      "827:\tlearn: 0.0002734\ttotal: 1m 1s\tremaining: 12.7s\n",
      "828:\tlearn: 0.0002731\ttotal: 1m 1s\tremaining: 12.6s\n",
      "829:\tlearn: 0.0002729\ttotal: 1m 1s\tremaining: 12.5s\n",
      "830:\tlearn: 0.0002723\ttotal: 1m 1s\tremaining: 12.5s\n",
      "831:\tlearn: 0.0002721\ttotal: 1m 1s\tremaining: 12.4s\n",
      "832:\tlearn: 0.0002720\ttotal: 1m 1s\tremaining: 12.3s\n",
      "833:\tlearn: 0.0002713\ttotal: 1m 1s\tremaining: 12.2s\n",
      "834:\tlearn: 0.0002701\ttotal: 1m 1s\tremaining: 12.2s\n",
      "835:\tlearn: 0.0002698\ttotal: 1m 1s\tremaining: 12.1s\n",
      "836:\tlearn: 0.0002696\ttotal: 1m 1s\tremaining: 12s\n",
      "837:\tlearn: 0.0002692\ttotal: 1m 1s\tremaining: 11.9s\n",
      "838:\tlearn: 0.0002681\ttotal: 1m 1s\tremaining: 11.9s\n",
      "839:\tlearn: 0.0002673\ttotal: 1m 1s\tremaining: 11.8s\n",
      "840:\tlearn: 0.0002649\ttotal: 1m 1s\tremaining: 11.7s\n",
      "841:\tlearn: 0.0002646\ttotal: 1m 1s\tremaining: 11.6s\n",
      "842:\tlearn: 0.0002644\ttotal: 1m 2s\tremaining: 11.6s\n",
      "843:\tlearn: 0.0002643\ttotal: 1m 2s\tremaining: 11.5s\n",
      "844:\tlearn: 0.0002636\ttotal: 1m 2s\tremaining: 11.4s\n",
      "845:\tlearn: 0.0002626\ttotal: 1m 2s\tremaining: 11.3s\n",
      "846:\tlearn: 0.0002623\ttotal: 1m 2s\tremaining: 11.3s\n",
      "847:\tlearn: 0.0002616\ttotal: 1m 2s\tremaining: 11.2s\n",
      "848:\tlearn: 0.0002605\ttotal: 1m 2s\tremaining: 11.1s\n",
      "849:\tlearn: 0.0002603\ttotal: 1m 2s\tremaining: 11s\n",
      "850:\tlearn: 0.0002601\ttotal: 1m 2s\tremaining: 11s\n",
      "851:\tlearn: 0.0002574\ttotal: 1m 2s\tremaining: 10.9s\n",
      "852:\tlearn: 0.0002571\ttotal: 1m 2s\tremaining: 10.8s\n",
      "853:\tlearn: 0.0002559\ttotal: 1m 2s\tremaining: 10.7s\n",
      "854:\tlearn: 0.0002557\ttotal: 1m 2s\tremaining: 10.7s\n",
      "855:\tlearn: 0.0002556\ttotal: 1m 2s\tremaining: 10.6s\n",
      "856:\tlearn: 0.0002554\ttotal: 1m 3s\tremaining: 10.5s\n",
      "857:\tlearn: 0.0002542\ttotal: 1m 3s\tremaining: 10.4s\n",
      "858:\tlearn: 0.0002540\ttotal: 1m 3s\tremaining: 10.4s\n",
      "859:\tlearn: 0.0002539\ttotal: 1m 3s\tremaining: 10.3s\n",
      "860:\tlearn: 0.0002533\ttotal: 1m 3s\tremaining: 10.2s\n",
      "861:\tlearn: 0.0002520\ttotal: 1m 3s\tremaining: 10.1s\n",
      "862:\tlearn: 0.0002510\ttotal: 1m 3s\tremaining: 10.1s\n",
      "863:\tlearn: 0.0002509\ttotal: 1m 3s\tremaining: 10s\n",
      "864:\tlearn: 0.0002506\ttotal: 1m 3s\tremaining: 9.93s\n",
      "865:\tlearn: 0.0002500\ttotal: 1m 3s\tremaining: 9.85s\n",
      "866:\tlearn: 0.0002496\ttotal: 1m 3s\tremaining: 9.78s\n",
      "867:\tlearn: 0.0002496\ttotal: 1m 3s\tremaining: 9.7s\n",
      "868:\tlearn: 0.0002493\ttotal: 1m 3s\tremaining: 9.63s\n",
      "869:\tlearn: 0.0002492\ttotal: 1m 3s\tremaining: 9.55s\n",
      "870:\tlearn: 0.0002491\ttotal: 1m 4s\tremaining: 9.48s\n",
      "871:\tlearn: 0.0002472\ttotal: 1m 4s\tremaining: 9.41s\n",
      "872:\tlearn: 0.0002460\ttotal: 1m 4s\tremaining: 9.33s\n",
      "873:\tlearn: 0.0002442\ttotal: 1m 4s\tremaining: 9.26s\n",
      "874:\tlearn: 0.0002440\ttotal: 1m 4s\tremaining: 9.19s\n",
      "875:\tlearn: 0.0002438\ttotal: 1m 4s\tremaining: 9.11s\n",
      "876:\tlearn: 0.0002437\ttotal: 1m 4s\tremaining: 9.04s\n",
      "877:\tlearn: 0.0002436\ttotal: 1m 4s\tremaining: 8.96s\n",
      "878:\tlearn: 0.0002428\ttotal: 1m 4s\tremaining: 8.89s\n",
      "879:\tlearn: 0.0002421\ttotal: 1m 4s\tremaining: 8.81s\n",
      "880:\tlearn: 0.0002420\ttotal: 1m 4s\tremaining: 8.74s\n",
      "881:\tlearn: 0.0002417\ttotal: 1m 4s\tremaining: 8.66s\n",
      "882:\tlearn: 0.0002415\ttotal: 1m 4s\tremaining: 8.59s\n",
      "883:\tlearn: 0.0002413\ttotal: 1m 4s\tremaining: 8.52s\n",
      "884:\tlearn: 0.0002402\ttotal: 1m 4s\tremaining: 8.44s\n",
      "885:\tlearn: 0.0002400\ttotal: 1m 5s\tremaining: 8.37s\n",
      "886:\tlearn: 0.0002393\ttotal: 1m 5s\tremaining: 8.29s\n",
      "887:\tlearn: 0.0002391\ttotal: 1m 5s\tremaining: 8.22s\n",
      "888:\tlearn: 0.0002390\ttotal: 1m 5s\tremaining: 8.15s\n",
      "889:\tlearn: 0.0002388\ttotal: 1m 5s\tremaining: 8.07s\n",
      "890:\tlearn: 0.0002387\ttotal: 1m 5s\tremaining: 8s\n",
      "891:\tlearn: 0.0002377\ttotal: 1m 5s\tremaining: 7.92s\n",
      "892:\tlearn: 0.0002375\ttotal: 1m 5s\tremaining: 7.85s\n",
      "893:\tlearn: 0.0002373\ttotal: 1m 5s\tremaining: 7.78s\n",
      "894:\tlearn: 0.0002371\ttotal: 1m 5s\tremaining: 7.7s\n",
      "895:\tlearn: 0.0002362\ttotal: 1m 5s\tremaining: 7.63s\n",
      "896:\tlearn: 0.0002361\ttotal: 1m 5s\tremaining: 7.56s\n",
      "897:\tlearn: 0.0002358\ttotal: 1m 5s\tremaining: 7.48s\n",
      "898:\tlearn: 0.0002352\ttotal: 1m 5s\tremaining: 7.41s\n",
      "899:\tlearn: 0.0002351\ttotal: 1m 6s\tremaining: 7.33s\n",
      "900:\tlearn: 0.0002343\ttotal: 1m 6s\tremaining: 7.26s\n",
      "901:\tlearn: 0.0002335\ttotal: 1m 6s\tremaining: 7.19s\n",
      "902:\tlearn: 0.0002334\ttotal: 1m 6s\tremaining: 7.11s\n",
      "903:\tlearn: 0.0002334\ttotal: 1m 6s\tremaining: 7.04s\n",
      "904:\tlearn: 0.0002321\ttotal: 1m 6s\tremaining: 6.96s\n",
      "905:\tlearn: 0.0002315\ttotal: 1m 6s\tremaining: 6.89s\n",
      "906:\tlearn: 0.0002315\ttotal: 1m 6s\tremaining: 6.82s\n",
      "907:\tlearn: 0.0002311\ttotal: 1m 6s\tremaining: 6.74s\n",
      "908:\tlearn: 0.0002304\ttotal: 1m 6s\tremaining: 6.67s\n",
      "909:\tlearn: 0.0002303\ttotal: 1m 6s\tremaining: 6.59s\n",
      "910:\tlearn: 0.0002297\ttotal: 1m 6s\tremaining: 6.52s\n",
      "911:\tlearn: 0.0002285\ttotal: 1m 6s\tremaining: 6.45s\n",
      "912:\tlearn: 0.0002284\ttotal: 1m 6s\tremaining: 6.37s\n",
      "913:\tlearn: 0.0002283\ttotal: 1m 6s\tremaining: 6.3s\n",
      "914:\tlearn: 0.0002275\ttotal: 1m 7s\tremaining: 6.23s\n",
      "915:\tlearn: 0.0002273\ttotal: 1m 7s\tremaining: 6.15s\n",
      "916:\tlearn: 0.0002272\ttotal: 1m 7s\tremaining: 6.08s\n",
      "917:\tlearn: 0.0002265\ttotal: 1m 7s\tremaining: 6.01s\n",
      "918:\tlearn: 0.0002265\ttotal: 1m 7s\tremaining: 5.94s\n",
      "919:\tlearn: 0.0002264\ttotal: 1m 7s\tremaining: 5.87s\n",
      "920:\tlearn: 0.0002257\ttotal: 1m 7s\tremaining: 5.8s\n",
      "921:\tlearn: 0.0002256\ttotal: 1m 7s\tremaining: 5.74s\n",
      "922:\tlearn: 0.0002255\ttotal: 1m 7s\tremaining: 5.67s\n",
      "923:\tlearn: 0.0002247\ttotal: 1m 8s\tremaining: 5.59s\n",
      "924:\tlearn: 0.0002235\ttotal: 1m 8s\tremaining: 5.52s\n",
      "925:\tlearn: 0.0002228\ttotal: 1m 8s\tremaining: 5.45s\n",
      "926:\tlearn: 0.0002224\ttotal: 1m 8s\tremaining: 5.37s\n",
      "927:\tlearn: 0.0002224\ttotal: 1m 8s\tremaining: 5.3s\n",
      "928:\tlearn: 0.0002223\ttotal: 1m 8s\tremaining: 5.22s\n",
      "929:\tlearn: 0.0002214\ttotal: 1m 8s\tremaining: 5.15s\n",
      "930:\tlearn: 0.0002213\ttotal: 1m 8s\tremaining: 5.08s\n",
      "931:\tlearn: 0.0002212\ttotal: 1m 8s\tremaining: 5s\n",
      "932:\tlearn: 0.0002210\ttotal: 1m 8s\tremaining: 4.93s\n",
      "933:\tlearn: 0.0002199\ttotal: 1m 8s\tremaining: 4.85s\n",
      "934:\tlearn: 0.0002191\ttotal: 1m 8s\tremaining: 4.78s\n",
      "935:\tlearn: 0.0002177\ttotal: 1m 8s\tremaining: 4.71s\n",
      "936:\tlearn: 0.0002176\ttotal: 1m 8s\tremaining: 4.63s\n",
      "937:\tlearn: 0.0002175\ttotal: 1m 8s\tremaining: 4.56s\n",
      "938:\tlearn: 0.0002174\ttotal: 1m 9s\tremaining: 4.48s\n",
      "939:\tlearn: 0.0002173\ttotal: 1m 9s\tremaining: 4.41s\n",
      "940:\tlearn: 0.0002171\ttotal: 1m 9s\tremaining: 4.33s\n",
      "941:\tlearn: 0.0002156\ttotal: 1m 9s\tremaining: 4.26s\n",
      "942:\tlearn: 0.0002152\ttotal: 1m 9s\tremaining: 4.19s\n",
      "943:\tlearn: 0.0002151\ttotal: 1m 9s\tremaining: 4.11s\n",
      "944:\tlearn: 0.0002145\ttotal: 1m 9s\tremaining: 4.04s\n",
      "945:\tlearn: 0.0002142\ttotal: 1m 9s\tremaining: 3.97s\n",
      "946:\tlearn: 0.0002138\ttotal: 1m 9s\tremaining: 3.89s\n",
      "947:\tlearn: 0.0002137\ttotal: 1m 9s\tremaining: 3.82s\n",
      "948:\tlearn: 0.0002131\ttotal: 1m 9s\tremaining: 3.74s\n",
      "949:\tlearn: 0.0002125\ttotal: 1m 9s\tremaining: 3.67s\n",
      "950:\tlearn: 0.0002125\ttotal: 1m 9s\tremaining: 3.6s\n",
      "951:\tlearn: 0.0002123\ttotal: 1m 9s\tremaining: 3.52s\n",
      "952:\tlearn: 0.0002120\ttotal: 1m 9s\tremaining: 3.45s\n",
      "953:\tlearn: 0.0002118\ttotal: 1m 10s\tremaining: 3.38s\n",
      "954:\tlearn: 0.0002114\ttotal: 1m 10s\tremaining: 3.3s\n",
      "955:\tlearn: 0.0002113\ttotal: 1m 10s\tremaining: 3.23s\n",
      "956:\tlearn: 0.0002107\ttotal: 1m 10s\tremaining: 3.15s\n",
      "957:\tlearn: 0.0002106\ttotal: 1m 10s\tremaining: 3.08s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "958:\tlearn: 0.0002105\ttotal: 1m 10s\tremaining: 3.01s\n",
      "959:\tlearn: 0.0002101\ttotal: 1m 10s\tremaining: 2.93s\n",
      "960:\tlearn: 0.0002100\ttotal: 1m 10s\tremaining: 2.86s\n",
      "961:\tlearn: 0.0002095\ttotal: 1m 10s\tremaining: 2.79s\n",
      "962:\tlearn: 0.0002092\ttotal: 1m 10s\tremaining: 2.71s\n",
      "963:\tlearn: 0.0002091\ttotal: 1m 10s\tremaining: 2.64s\n",
      "964:\tlearn: 0.0002086\ttotal: 1m 10s\tremaining: 2.57s\n",
      "965:\tlearn: 0.0002082\ttotal: 1m 10s\tremaining: 2.49s\n",
      "966:\tlearn: 0.0002080\ttotal: 1m 10s\tremaining: 2.42s\n",
      "967:\tlearn: 0.0002077\ttotal: 1m 10s\tremaining: 2.35s\n",
      "968:\tlearn: 0.0002073\ttotal: 1m 11s\tremaining: 2.27s\n",
      "969:\tlearn: 0.0002072\ttotal: 1m 11s\tremaining: 2.2s\n",
      "970:\tlearn: 0.0002058\ttotal: 1m 11s\tremaining: 2.13s\n",
      "971:\tlearn: 0.0002057\ttotal: 1m 11s\tremaining: 2.05s\n",
      "972:\tlearn: 0.0002056\ttotal: 1m 11s\tremaining: 1.98s\n",
      "973:\tlearn: 0.0002053\ttotal: 1m 11s\tremaining: 1.9s\n",
      "974:\tlearn: 0.0002052\ttotal: 1m 11s\tremaining: 1.83s\n",
      "975:\tlearn: 0.0002051\ttotal: 1m 11s\tremaining: 1.76s\n",
      "976:\tlearn: 0.0002049\ttotal: 1m 11s\tremaining: 1.68s\n",
      "977:\tlearn: 0.0002047\ttotal: 1m 11s\tremaining: 1.61s\n",
      "978:\tlearn: 0.0002046\ttotal: 1m 11s\tremaining: 1.54s\n",
      "979:\tlearn: 0.0002043\ttotal: 1m 11s\tremaining: 1.46s\n",
      "980:\tlearn: 0.0002034\ttotal: 1m 11s\tremaining: 1.39s\n",
      "981:\tlearn: 0.0002022\ttotal: 1m 11s\tremaining: 1.32s\n",
      "982:\tlearn: 0.0002019\ttotal: 1m 11s\tremaining: 1.24s\n",
      "983:\tlearn: 0.0002017\ttotal: 1m 12s\tremaining: 1.17s\n",
      "984:\tlearn: 0.0002007\ttotal: 1m 12s\tremaining: 1.1s\n",
      "985:\tlearn: 0.0002003\ttotal: 1m 12s\tremaining: 1.02s\n",
      "986:\tlearn: 0.0002001\ttotal: 1m 12s\tremaining: 952ms\n",
      "987:\tlearn: 0.0001997\ttotal: 1m 12s\tremaining: 878ms\n",
      "988:\tlearn: 0.0001992\ttotal: 1m 12s\tremaining: 805ms\n",
      "989:\tlearn: 0.0001990\ttotal: 1m 12s\tremaining: 732ms\n",
      "990:\tlearn: 0.0001985\ttotal: 1m 12s\tremaining: 659ms\n",
      "991:\tlearn: 0.0001984\ttotal: 1m 12s\tremaining: 585ms\n",
      "992:\tlearn: 0.0001975\ttotal: 1m 12s\tremaining: 512ms\n",
      "993:\tlearn: 0.0001969\ttotal: 1m 12s\tremaining: 439ms\n",
      "994:\tlearn: 0.0001968\ttotal: 1m 12s\tremaining: 366ms\n",
      "995:\tlearn: 0.0001967\ttotal: 1m 12s\tremaining: 293ms\n",
      "996:\tlearn: 0.0001964\ttotal: 1m 12s\tremaining: 219ms\n",
      "997:\tlearn: 0.0001957\ttotal: 1m 12s\tremaining: 146ms\n",
      "998:\tlearn: 0.0001950\ttotal: 1m 13s\tremaining: 73.1ms\n",
      "999:\tlearn: 0.0001939\ttotal: 1m 13s\tremaining: 0us\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<catboost.core.CatBoostClassifier at 0x7fed7a55c278>"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from catboost import CatBoostClassifier\n",
    "\n",
    "cb_clf = CatBoostClassifier()\n",
    "cb_clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Result:\n",
      "================================================\n",
      "Accuracy Score: 100.00%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                  0      1  accuracy  macro avg  weighted avg\n",
      "precision      1.00   1.00      1.00       1.00          1.00\n",
      "recall         1.00   0.98      1.00       0.99          1.00\n",
      "f1-score       1.00   0.99      1.00       1.00          1.00\n",
      "support   159204.00 287.00      1.00  159491.00     159491.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[159204      0]\n",
      " [     5    282]]\n",
      "\n",
      "Test Result:\n",
      "================================================\n",
      "Accuracy Score: 99.97%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                 0      1  accuracy  macro avg  weighted avg\n",
      "precision     1.00   0.96      1.00       0.98          1.00\n",
      "recall        1.00   0.82      1.00       0.91          1.00\n",
      "f1-score      1.00   0.89      1.00       0.94          1.00\n",
      "support   85307.00 136.00      1.00   85443.00      85443.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[85302     5]\n",
      " [   24   112]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "y_train_pred = cb_clf.predict(X_train)\n",
    "y_test_pred = cb_clf.predict(X_test)\n",
    "\n",
    "print_score(y_train, y_train_pred, train=True)\n",
    "print_score(y_test, y_test_pred, train=False)\n",
    "\n",
    "scores_dict['CatBoost'] = {\n",
    "        'Train': f1_score(y_train,y_train_pred),\n",
    "        'Test': f1_score(y_test, y_test_pred),\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. 5. LigthGBM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Result:\n",
      "================================================\n",
      "Accuracy Score: 99.86%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                  0      1  accuracy  macro avg  weighted avg\n",
      "precision      1.00   0.59      1.00       0.80          1.00\n",
      "recall         1.00   0.69      1.00       0.84          1.00\n",
      "f1-score       1.00   0.64      1.00       0.82          1.00\n",
      "support   159204.00 287.00      1.00  159491.00     159491.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[159068    136]\n",
      " [    89    198]]\n",
      "\n",
      "Test Result:\n",
      "================================================\n",
      "Accuracy Score: 99.75%\n",
      "_______________________________________________\n",
      "Classification Report:\n",
      "                 0      1  accuracy  macro avg  weighted avg\n",
      "precision     1.00   0.35      1.00       0.67          1.00\n",
      "recall        1.00   0.66      1.00       0.83          1.00\n",
      "f1-score      1.00   0.46      1.00       0.73          1.00\n",
      "support   85307.00 136.00      1.00   85443.00      85443.00\n",
      "_______________________________________________\n",
      "Confusion Matrix: \n",
      " [[85140   167]\n",
      " [   46    90]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from lightgbm import LGBMClassifier\n",
    "\n",
    "lgbm_clf = LGBMClassifier()\n",
    "lgbm_clf.fit(X_train, y_train)\n",
    "\n",
    "y_train_pred = lgbm_clf.predict(X_train)\n",
    "y_test_pred = lgbm_clf.predict(X_test)\n",
    "\n",
    "print_score(y_train, y_train_pred, train=True)\n",
    "print_score(y_test, y_test_pred, train=False)\n",
    "\n",
    "scores_dict['LigthGBM'] = {\n",
    "        'Train': f1_score(y_train,y_train_pred),\n",
    "        'Test': f1_score(y_test, y_test_pred),\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5. Model Comparaison"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fed7a662c50>"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "scores_df = pd.DataFrame(scores_dict)\n",
    "\n",
    "scores_df.plot(kind='barh', figsize=(15, 8))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}