LSTM_2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "1fac16a0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-08-26 17:15:07.385480: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\n",
      "To enable the following instructions: AVX2 AVX512F FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2023-08-26 17:15:08.256909: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, LSTM, Dropout, Flatten, Input\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import cross_val_predict, cross_val_score\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import accuracy_score,confusion_matrix, roc_curve, roc_auc_score, auc, classification_report\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score\n",
    "\n",
    "\n",
    "from keras.layers import Conv2D, MaxPooling2D\n",
    "from keras.layers import Conv2D, Conv1D, MaxPooling2D, MaxPooling1D, BatchNormalization\n",
    "\n",
    "from keras.utils import plot_model\n",
    "from keras import backend as K\n",
    "import keras\n",
    "\n",
    "%matplotlib  inline\n",
    "\n",
    "#random seed\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2cb5eef2",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"writeData.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c9afdcba",
   "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>Unnamed: 0</th>\n",
       "      <th>78</th>\n",
       "      <th>235</th>\n",
       "      <th>4.1</th>\n",
       "      <th>22</th>\n",
       "      <th>2</th>\n",
       "      <th>219</th>\n",
       "      <th>0</th>\n",
       "      <th>22.1</th>\n",
       "      <th>69</th>\n",
       "      <th>287</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>264</td>\n",
       "      <td>1468</td>\n",
       "      <td>23</td>\n",
       "      <td>22</td>\n",
       "      <td>10</td>\n",
       "      <td>581</td>\n",
       "      <td>1</td>\n",
       "      <td>22</td>\n",
       "      <td>255</td>\n",
       "      <td>338</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>94</td>\n",
       "      <td>192</td>\n",
       "      <td>16</td>\n",
       "      <td>12</td>\n",
       "      <td>11</td>\n",
       "      <td>141</td>\n",
       "      <td>3</td>\n",
       "      <td>12</td>\n",
       "      <td>93</td>\n",
       "      <td>247</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>124</td>\n",
       "      <td>434</td>\n",
       "      <td>136</td>\n",
       "      <td>16</td>\n",
       "      <td>4</td>\n",
       "      <td>234</td>\n",
       "      <td>21</td>\n",
       "      <td>16</td>\n",
       "      <td>105</td>\n",
       "      <td>318</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>105</td>\n",
       "      <td>251</td>\n",
       "      <td>34</td>\n",
       "      <td>14</td>\n",
       "      <td>16</td>\n",
       "      <td>301</td>\n",
       "      <td>9</td>\n",
       "      <td>14</td>\n",
       "      <td>89</td>\n",
       "      <td>263</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>80</td>\n",
       "      <td>374</td>\n",
       "      <td>10</td>\n",
       "      <td>11</td>\n",
       "      <td>106</td>\n",
       "      <td>233</td>\n",
       "      <td>3</td>\n",
       "      <td>11</td>\n",
       "      <td>76</td>\n",
       "      <td>305</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0   78   235  4.1  22    2  219   0  22.1   69  287  1\n",
       "0           0  264  1468   23  22   10  581   1    22  255  338  1\n",
       "1           1   94   192   16  12   11  141   3    12   93  247  1\n",
       "2           2  124   434  136  16    4  234  21    16  105  318  1\n",
       "3           3  105   251   34  14   16  301   9    14   89  263  1\n",
       "4           4   80   374   10  11  106  233   3    11   76  305  1"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f2b2ae96",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of dataset:  (7199, 12)\n"
     ]
    }
   ],
   "source": [
    "print(\"Shape of dataset: \",df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2bd6a958",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Unnamed: 0', '78', '235', '4.1', '22', '2', '219', '0', '22.1', '69',\n",
       "       '287', '1'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "66e07d17",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1, -1])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['1'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "01eed47f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# checking null values\n",
    "df.isnull().sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "6c32e18a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1, -1])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# after removing nan values\n",
    "df['1'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "841f7ed0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1\n",
       " 1    7064\n",
       "-1     135\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['1'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3a53ab58",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1\n",
       "-1    7064\n",
       " 1     135\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['1'] = df['1']*-1\n",
    "df['1'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "77600aa0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "length of dataset:  7199\n"
     ]
    }
   ],
   "source": [
    "print(\"length of dataset: \", len(df))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "95a03414",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Unnamed: 0    0\n",
       "78            0\n",
       "235           0\n",
       "4.1           0\n",
       "22            0\n",
       "2             0\n",
       "219           0\n",
       "0             0\n",
       "22.1          0\n",
       "69            0\n",
       "287           0\n",
       "1             0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# checking null values\n",
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "806d3a7a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Unnamed: 0    int64\n",
       "78            int64\n",
       "235           int64\n",
       "4.1           int64\n",
       "22            int64\n",
       "2             int64\n",
       "219           int64\n",
       "0             int64\n",
       "22.1          int64\n",
       "69            int64\n",
       "287           int64\n",
       "1             int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "155c6c0d",
   "metadata": {},
   "outputs": [],
   "source": [
    "y = df['1'].values\n",
    "X = df.drop(columns = [\"Unnamed: 0\", \"1\"]).values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "3b50c875",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((7199, 10), (7199,))"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape, y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "137de4ed",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Noramlize X values in the range (0,1)\n",
    "sc = MinMaxScaler(feature_range = (0, 1))\n",
    "X_scaled = sc.fit_transform(X)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "3eed84e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(7199, 10)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_scaled.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "98633314",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "num_classes = len(np.unique(y))\n",
    "num_classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "c19a898d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reshape the data for LSTM (samples, timesteps, features)\n",
    "sequence_length = 10 # Number of timesteps\n",
    "X_seq = np.array([X_scaled[i:i+sequence_length, :] for i in range(len(X_scaled) - sequence_length + 1)])\n",
    "y_seq = y[sequence_length - 1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "fe72198b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((7190, 10, 10), (7190,))"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_seq.shape, y_seq.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d284d1b1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# train test split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_seq, y_seq, test_size=0.15, random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "d0a267e3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of X_train:  (6111, 10, 10)\n",
      "Shape of X_test:  (1079, 10, 10)\n",
      "Shape of y_train:  (6111,)\n",
      "Shape of y_test:  (1079,)\n"
     ]
    }
   ],
   "source": [
    "print(\"Shape of X_train: \", X_train.shape)\n",
    "print(\"Shape of X_test: \", X_test.shape)\n",
    "print(\"Shape of y_train: \", y_train.shape)\n",
    "print(\"Shape of y_test: \", y_test.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "8642723d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of X_train:  (6111, 10, 10)\n",
      "Shape of X_test:  (1079, 10, 10)\n",
      "Shape of y_train:  (6111,)\n",
      "Shape of y_test:  (1079,)\n"
     ]
    }
   ],
   "source": [
    "# After reshaping\n",
    "print(\"Shape of X_train: \", X_train.shape)\n",
    "print(\"Shape of X_test: \", X_test.shape)\n",
    "print(\"Shape of y_train: \", y_train.shape)\n",
    "print(\"Shape of y_test: \", y_test.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "bad5afd9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Now we shuffle the training set because we will be using the validation_split option later when training.\n",
    "idx = np.random.permutation(len(X_train))\n",
    "X_train = X_train[idx]\n",
    "y_train = y_train[idx]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "c0c193a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Standardize the labels to positive integers. The expected labels will then be 0 and 1.\n",
    "y_train[y_train == -1] = 0\n",
    "y_test[y_test == -1] = 0\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "a7791683",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((10, 10), 10)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape[1:], X_train.shape[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3f5868f0",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "db8780d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10, 10)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "input_shape = X_train.shape[1:]\n",
    "input_shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "5a87b413",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-08-26 17:15:13.519501: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.587642: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.587932: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.590168: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.590415: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.590598: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.705563: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.705836: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.706021: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:996] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero. See more at https://github.com/torvalds/linux/blob/v6.0/Documentation/ABI/testing/sysfs-bus-pci#L344-L355\n",
      "2023-08-26 17:15:13.706210: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1635] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 10252 MB memory:  -> device: 0, name: NVIDIA GeForce RTX 3080 Ti, pci bus id: 0000:00:10.0, compute capability: 8.6\n",
      "2023-08-26 17:15:14.303506: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_2_grad/concat/split_2/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_2_grad/concat/split_2/split_dim}}]]\n",
      "2023-08-26 17:15:14.304831: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_grad/concat/split/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_grad/concat/split/split_dim}}]]\n",
      "2023-08-26 17:15:14.307281: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_1_grad/concat/split_1/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_1_grad/concat/split_1/split_dim}}]]\n",
      "2023-08-26 17:15:14.537439: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_2_grad/concat/split_2/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_2_grad/concat/split_2/split_dim}}]]\n",
      "2023-08-26 17:15:14.539172: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_grad/concat/split/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_grad/concat/split/split_dim}}]]\n",
      "2023-08-26 17:15:14.541121: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_1_grad/concat/split_1/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_1_grad/concat/split_1/split_dim}}]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create an LSTM model with two hidden layers and dropout\n",
    "model = Sequential()\n",
    "model.add(LSTM(64, input_shape=(X_train.shape[1], X_train.shape[2]), return_sequences=True))\n",
    "model.add(Dropout(0.2))  # Dropout for the first LSTM layer\n",
    "\n",
    "model.add(LSTM(64))\n",
    "model.add(Dropout(0.2))  # Dropout for the second LSTM layer\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "\n",
    "plot_model(model, to_file=\"LSTM_model_feature_selection.png\", show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "0b30f2c7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm (LSTM)                 (None, 10, 64)            19200     \n",
      "                                                                 \n",
      " dropout (Dropout)           (None, 10, 64)            0         \n",
      "                                                                 \n",
      " lstm_1 (LSTM)               (None, 64)                33024     \n",
      "                                                                 \n",
      " dropout_1 (Dropout)         (None, 64)                0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 1)                 65        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 52,289\n",
      "Trainable params: 52,289\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8184fa15",
   "metadata": {},
   "source": [
    "# Train the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "289069a9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/500\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-08-26 17:15:15.341080: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_2_grad/concat/split_2/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_2_grad/concat/split_2/split_dim}}]]\n",
      "2023-08-26 17:15:15.342774: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_grad/concat/split/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_grad/concat/split/split_dim}}]]\n",
      "2023-08-26 17:15:15.344660: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_1_grad/concat/split_1/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_1_grad/concat/split_1/split_dim}}]]\n",
      "2023-08-26 17:15:15.566551: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_2_grad/concat/split_2/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_2_grad/concat/split_2/split_dim}}]]\n",
      "2023-08-26 17:15:15.568509: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_grad/concat/split/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_grad/concat/split/split_dim}}]]\n",
      "2023-08-26 17:15:15.570233: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_1_grad/concat/split_1/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_1_grad/concat/split_1/split_dim}}]]\n",
      "2023-08-26 17:15:16.630735: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_2_grad/concat/split_2/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_2_grad/concat/split_2/split_dim}}]]\n",
      "2023-08-26 17:15:16.633361: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_grad/concat/split/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_grad/concat/split/split_dim}}]]\n",
      "2023-08-26 17:15:16.634984: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_1_grad/concat/split_1/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_1_grad/concat/split_1/split_dim}}]]\n",
      "2023-08-26 17:15:16.851866: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_2_grad/concat/split_2/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_2_grad/concat/split_2/split_dim}}]]\n",
      "2023-08-26 17:15:16.853714: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_grad/concat/split/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_grad/concat/split/split_dim}}]]\n",
      "2023-08-26 17:15:16.855457: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_1_grad/concat/split_1/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_1_grad/concat/split_1/split_dim}}]]\n",
      "2023-08-26 17:15:17.944231: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:424] Loaded cuDNN version 8600\n",
      "2023-08-26 17:15:18.056532: I tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:637] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n",
      "2023-08-26 17:15:18.067691: I tensorflow/compiler/xla/service/service.cc:169] XLA service 0x940c830 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n",
      "2023-08-26 17:15:18.067731: I tensorflow/compiler/xla/service/service.cc:177]   StreamExecutor device (0): NVIDIA GeForce RTX 3080 Ti, Compute Capability 8.6\n",
      "2023-08-26 17:15:18.075120: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:269] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n",
      "2023-08-26 17:15:18.306057: I ./tensorflow/compiler/jit/device_compiler.h:180] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "151/153 [============================>.] - ETA: 0s - loss: 0.0324 - accuracy: 0.9770"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-08-26 17:15:21.017442: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_2_grad/concat/split_2/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_2_grad/concat/split_2/split_dim}}]]\n",
      "2023-08-26 17:15:21.019610: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_grad/concat/split/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_grad/concat/split/split_dim}}]]\n",
      "2023-08-26 17:15:21.021461: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_1_grad/concat/split_1/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_1_grad/concat/split_1/split_dim}}]]\n",
      "2023-08-26 17:15:21.218368: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_2_grad/concat/split_2/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_2_grad/concat/split_2/split_dim}}]]\n",
      "2023-08-26 17:15:21.219817: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_grad/concat/split/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_grad/concat/split/split_dim}}]]\n",
      "2023-08-26 17:15:21.221876: I tensorflow/core/common_runtime/executor.cc:1197] [/device:CPU:0] (DEBUG INFO) Executor start aborting (this does not indicate an error and you can ignore this message): INVALID_ARGUMENT: You must feed a value for placeholder tensor 'gradients/split_1_grad/concat/split_1/split_dim' with dtype int32\n",
      "\t [[{{node gradients/split_1_grad/concat/split_1/split_dim}}]]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 7s 19ms/step - loss: 0.0321 - accuracy: 0.9773 - val_loss: 0.0180 - val_accuracy: 0.9820\n",
      "Epoch 2/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0184 - accuracy: 0.9816 - val_loss: 0.0178 - val_accuracy: 0.9820\n",
      "Epoch 3/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0181 - accuracy: 0.9816 - val_loss: 0.0176 - val_accuracy: 0.9820\n",
      "Epoch 4/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0176 - val_accuracy: 0.9820\n",
      "Epoch 5/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 6/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0176 - val_accuracy: 0.9820\n",
      "Epoch 7/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0181 - accuracy: 0.9816 - val_loss: 0.0178 - val_accuracy: 0.9820\n",
      "Epoch 8/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0176 - val_accuracy: 0.9820\n",
      "Epoch 9/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 10/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 11/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0179 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 12/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 13/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0177 - val_accuracy: 0.9820\n",
      "Epoch 14/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0176 - val_accuracy: 0.9820\n",
      "Epoch 15/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 16/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 17/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 18/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0180 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 19/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 20/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0179 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 21/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 22/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0179 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 23/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0176 - val_accuracy: 0.9820\n",
      "Epoch 24/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0179 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 25/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0176 - val_accuracy: 0.9820\n",
      "Epoch 26/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 27/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 28/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0179 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 29/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 30/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0179 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 31/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 32/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 33/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 34/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 35/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 36/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 37/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 38/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 39/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 40/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 41/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 42/500\n",
      "153/153 [==============================] - 2s 12ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0174 - val_accuracy: 0.9820\n",
      "Epoch 43/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 44/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 45/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 46/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 47/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0175 - val_accuracy: 0.9820\n",
      "Epoch 48/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 49/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 50/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 51/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 52/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 53/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 54/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 55/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 56/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 57/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 58/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 59/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 60/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 61/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 62/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 63/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 64/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0170 - val_accuracy: 0.9820\n",
      "Epoch 65/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 66/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 67/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0175 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 68/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0177 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 69/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0170 - val_accuracy: 0.9820\n",
      "Epoch 70/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0176 - accuracy: 0.9816 - val_loss: 0.0170 - val_accuracy: 0.9820\n",
      "Epoch 71/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0175 - accuracy: 0.9816 - val_loss: 0.0170 - val_accuracy: 0.9820\n",
      "Epoch 72/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0175 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 73/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0178 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 74/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0174 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 75/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0175 - accuracy: 0.9816 - val_loss: 0.0169 - val_accuracy: 0.9820\n",
      "Epoch 76/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0175 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 77/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0174 - accuracy: 0.9816 - val_loss: 0.0170 - val_accuracy: 0.9820\n",
      "Epoch 78/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0173 - accuracy: 0.9816 - val_loss: 0.0176 - val_accuracy: 0.9820\n",
      "Epoch 79/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0174 - accuracy: 0.9816 - val_loss: 0.0168 - val_accuracy: 0.9820\n",
      "Epoch 80/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0172 - accuracy: 0.9816 - val_loss: 0.0168 - val_accuracy: 0.9820\n",
      "Epoch 81/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0175 - accuracy: 0.9816 - val_loss: 0.0169 - val_accuracy: 0.9820\n",
      "Epoch 82/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0171 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 83/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0171 - accuracy: 0.9816 - val_loss: 0.0170 - val_accuracy: 0.9820\n",
      "Epoch 84/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0170 - accuracy: 0.9816 - val_loss: 0.0168 - val_accuracy: 0.9820\n",
      "Epoch 85/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0173 - accuracy: 0.9816 - val_loss: 0.0167 - val_accuracy: 0.9820\n",
      "Epoch 86/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0170 - accuracy: 0.9816 - val_loss: 0.0173 - val_accuracy: 0.9820\n",
      "Epoch 87/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0169 - accuracy: 0.9816 - val_loss: 0.0167 - val_accuracy: 0.9820\n",
      "Epoch 88/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0168 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 89/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0168 - accuracy: 0.9816 - val_loss: 0.0165 - val_accuracy: 0.9820\n",
      "Epoch 90/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0169 - accuracy: 0.9816 - val_loss: 0.0167 - val_accuracy: 0.9820\n",
      "Epoch 91/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0170 - accuracy: 0.9816 - val_loss: 0.0169 - val_accuracy: 0.9820\n",
      "Epoch 92/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0169 - accuracy: 0.9816 - val_loss: 0.0167 - val_accuracy: 0.9820\n",
      "Epoch 93/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0168 - accuracy: 0.9816 - val_loss: 0.0168 - val_accuracy: 0.9820\n",
      "Epoch 94/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0164 - accuracy: 0.9816 - val_loss: 0.0169 - val_accuracy: 0.9820\n",
      "Epoch 95/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0165 - accuracy: 0.9816 - val_loss: 0.0165 - val_accuracy: 0.9820\n",
      "Epoch 96/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0162 - accuracy: 0.9816 - val_loss: 0.0171 - val_accuracy: 0.9820\n",
      "Epoch 97/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0164 - accuracy: 0.9816 - val_loss: 0.0172 - val_accuracy: 0.9820\n",
      "Epoch 98/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0162 - accuracy: 0.9816 - val_loss: 0.0165 - val_accuracy: 0.9820\n",
      "Epoch 99/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0161 - accuracy: 0.9814 - val_loss: 0.0161 - val_accuracy: 0.9820\n",
      "Epoch 100/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0162 - accuracy: 0.9816 - val_loss: 0.0169 - val_accuracy: 0.9820\n",
      "Epoch 101/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0157 - accuracy: 0.9820 - val_loss: 0.0170 - val_accuracy: 0.9820\n",
      "Epoch 102/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0162 - accuracy: 0.9816 - val_loss: 0.0162 - val_accuracy: 0.9820\n",
      "Epoch 103/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0157 - accuracy: 0.9818 - val_loss: 0.0169 - val_accuracy: 0.9820\n",
      "Epoch 104/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0158 - accuracy: 0.9816 - val_loss: 0.0161 - val_accuracy: 0.9820\n",
      "Epoch 105/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0153 - accuracy: 0.9818 - val_loss: 0.0170 - val_accuracy: 0.9787\n",
      "Epoch 106/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0172 - accuracy: 0.9810 - val_loss: 0.0172 - val_accuracy: 0.9812\n",
      "Epoch 107/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0151 - accuracy: 0.9826 - val_loss: 0.0164 - val_accuracy: 0.9820\n",
      "Epoch 108/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0156 - accuracy: 0.9820 - val_loss: 0.0180 - val_accuracy: 0.9820\n",
      "Epoch 109/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0167 - accuracy: 0.9824 - val_loss: 0.0158 - val_accuracy: 0.9820\n",
      "Epoch 110/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0149 - accuracy: 0.9820 - val_loss: 0.0164 - val_accuracy: 0.9812\n",
      "Epoch 111/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0149 - accuracy: 0.9824 - val_loss: 0.0172 - val_accuracy: 0.9812\n",
      "Epoch 112/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0151 - accuracy: 0.9832 - val_loss: 0.0166 - val_accuracy: 0.9820\n",
      "Epoch 113/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0154 - accuracy: 0.9824 - val_loss: 0.0162 - val_accuracy: 0.9820\n",
      "Epoch 114/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0147 - accuracy: 0.9834 - val_loss: 0.0168 - val_accuracy: 0.9836\n",
      "Epoch 115/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0149 - accuracy: 0.9830 - val_loss: 0.0154 - val_accuracy: 0.9828\n",
      "Epoch 116/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0139 - accuracy: 0.9849 - val_loss: 0.0179 - val_accuracy: 0.9820\n",
      "Epoch 117/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0152 - accuracy: 0.9836 - val_loss: 0.0165 - val_accuracy: 0.9828\n",
      "Epoch 118/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0145 - accuracy: 0.9830 - val_loss: 0.0152 - val_accuracy: 0.9836\n",
      "Epoch 119/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0138 - accuracy: 0.9838 - val_loss: 0.0172 - val_accuracy: 0.9812\n",
      "Epoch 120/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0141 - accuracy: 0.9840 - val_loss: 0.0151 - val_accuracy: 0.9820\n",
      "Epoch 121/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0139 - accuracy: 0.9838 - val_loss: 0.0150 - val_accuracy: 0.9820\n",
      "Epoch 122/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0138 - accuracy: 0.9847 - val_loss: 0.0155 - val_accuracy: 0.9804\n",
      "Epoch 123/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0130 - accuracy: 0.9857 - val_loss: 0.0169 - val_accuracy: 0.9828\n",
      "Epoch 124/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0131 - accuracy: 0.9855 - val_loss: 0.0152 - val_accuracy: 0.9828\n",
      "Epoch 125/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0169 - accuracy: 0.9824 - val_loss: 0.0179 - val_accuracy: 0.9820\n",
      "Epoch 126/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0185 - accuracy: 0.9814 - val_loss: 0.0177 - val_accuracy: 0.9820\n",
      "Epoch 127/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0168 - accuracy: 0.9816 - val_loss: 0.0147 - val_accuracy: 0.9820\n",
      "Epoch 128/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0154 - accuracy: 0.9822 - val_loss: 0.0177 - val_accuracy: 0.9779\n",
      "Epoch 129/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0137 - accuracy: 0.9840 - val_loss: 0.0160 - val_accuracy: 0.9820\n",
      "Epoch 130/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0147 - accuracy: 0.9834 - val_loss: 0.0147 - val_accuracy: 0.9828\n",
      "Epoch 131/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0128 - accuracy: 0.9849 - val_loss: 0.0144 - val_accuracy: 0.9828\n",
      "Epoch 132/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0127 - accuracy: 0.9857 - val_loss: 0.0177 - val_accuracy: 0.9804\n",
      "Epoch 133/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0122 - accuracy: 0.9867 - val_loss: 0.0151 - val_accuracy: 0.9828\n",
      "Epoch 134/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0130 - accuracy: 0.9865 - val_loss: 0.0193 - val_accuracy: 0.9771\n",
      "Epoch 135/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0127 - accuracy: 0.9851 - val_loss: 0.0146 - val_accuracy: 0.9845\n",
      "Epoch 136/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0130 - accuracy: 0.9857 - val_loss: 0.0140 - val_accuracy: 0.9853\n",
      "Epoch 137/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0123 - accuracy: 0.9861 - val_loss: 0.0134 - val_accuracy: 0.9861\n",
      "Epoch 138/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0118 - accuracy: 0.9871 - val_loss: 0.0145 - val_accuracy: 0.9836\n",
      "Epoch 139/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0130 - accuracy: 0.9847 - val_loss: 0.0143 - val_accuracy: 0.9828\n",
      "Epoch 140/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0120 - accuracy: 0.9859 - val_loss: 0.0160 - val_accuracy: 0.9796\n",
      "Epoch 141/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0129 - accuracy: 0.9859 - val_loss: 0.0146 - val_accuracy: 0.9845\n",
      "Epoch 142/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0116 - accuracy: 0.9865 - val_loss: 0.0139 - val_accuracy: 0.9820\n",
      "Epoch 143/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0111 - accuracy: 0.9871 - val_loss: 0.0147 - val_accuracy: 0.9845\n",
      "Epoch 144/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0115 - accuracy: 0.9869 - val_loss: 0.0155 - val_accuracy: 0.9836\n",
      "Epoch 145/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0115 - accuracy: 0.9869 - val_loss: 0.0139 - val_accuracy: 0.9836\n",
      "Epoch 146/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0116 - accuracy: 0.9877 - val_loss: 0.0145 - val_accuracy: 0.9820\n",
      "Epoch 147/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0113 - accuracy: 0.9875 - val_loss: 0.0142 - val_accuracy: 0.9828\n",
      "Epoch 148/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0115 - accuracy: 0.9875 - val_loss: 0.0144 - val_accuracy: 0.9828\n",
      "Epoch 149/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0115 - accuracy: 0.9871 - val_loss: 0.0138 - val_accuracy: 0.9836\n",
      "Epoch 150/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0113 - accuracy: 0.9875 - val_loss: 0.0156 - val_accuracy: 0.9828\n",
      "Epoch 151/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0108 - accuracy: 0.9881 - val_loss: 0.0141 - val_accuracy: 0.9845\n",
      "Epoch 152/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0104 - accuracy: 0.9890 - val_loss: 0.0154 - val_accuracy: 0.9828\n",
      "Epoch 153/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0120 - accuracy: 0.9865 - val_loss: 0.0134 - val_accuracy: 0.9845\n",
      "Epoch 154/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0111 - accuracy: 0.9883 - val_loss: 0.0127 - val_accuracy: 0.9861\n",
      "Epoch 155/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0111 - accuracy: 0.9875 - val_loss: 0.0122 - val_accuracy: 0.9869\n",
      "Epoch 156/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0104 - accuracy: 0.9885 - val_loss: 0.0133 - val_accuracy: 0.9853\n",
      "Epoch 157/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0104 - accuracy: 0.9877 - val_loss: 0.0157 - val_accuracy: 0.9812\n",
      "Epoch 158/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0104 - accuracy: 0.9881 - val_loss: 0.0163 - val_accuracy: 0.9812\n",
      "Epoch 159/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0103 - accuracy: 0.9890 - val_loss: 0.0141 - val_accuracy: 0.9836\n",
      "Epoch 160/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0096 - accuracy: 0.9898 - val_loss: 0.0132 - val_accuracy: 0.9853\n",
      "Epoch 161/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0095 - accuracy: 0.9902 - val_loss: 0.0122 - val_accuracy: 0.9861\n",
      "Epoch 162/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0108 - accuracy: 0.9881 - val_loss: 0.0158 - val_accuracy: 0.9828\n",
      "Epoch 163/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0102 - accuracy: 0.9887 - val_loss: 0.0156 - val_accuracy: 0.9820\n",
      "Epoch 164/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0099 - accuracy: 0.9890 - val_loss: 0.0141 - val_accuracy: 0.9853\n",
      "Epoch 165/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0108 - accuracy: 0.9881 - val_loss: 0.0163 - val_accuracy: 0.9812\n",
      "Epoch 166/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0105 - accuracy: 0.9881 - val_loss: 0.0132 - val_accuracy: 0.9853\n",
      "Epoch 167/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0097 - accuracy: 0.9898 - val_loss: 0.0132 - val_accuracy: 0.9861\n",
      "Epoch 168/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0105 - accuracy: 0.9883 - val_loss: 0.0138 - val_accuracy: 0.9845\n",
      "Epoch 169/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0108 - accuracy: 0.9879 - val_loss: 0.0128 - val_accuracy: 0.9853\n",
      "Epoch 170/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0103 - accuracy: 0.9879 - val_loss: 0.0146 - val_accuracy: 0.9828\n",
      "Epoch 171/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0098 - accuracy: 0.9896 - val_loss: 0.0131 - val_accuracy: 0.9861\n",
      "Epoch 172/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0092 - accuracy: 0.9902 - val_loss: 0.0134 - val_accuracy: 0.9845\n",
      "Epoch 173/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0087 - accuracy: 0.9910 - val_loss: 0.0123 - val_accuracy: 0.9869\n",
      "Epoch 174/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0103 - accuracy: 0.9885 - val_loss: 0.0144 - val_accuracy: 0.9845\n",
      "Epoch 175/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0097 - accuracy: 0.9900 - val_loss: 0.0138 - val_accuracy: 0.9845\n",
      "Epoch 176/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0091 - accuracy: 0.9902 - val_loss: 0.0109 - val_accuracy: 0.9886\n",
      "Epoch 177/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0088 - accuracy: 0.9908 - val_loss: 0.0111 - val_accuracy: 0.9877\n",
      "Epoch 178/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0085 - accuracy: 0.9912 - val_loss: 0.0123 - val_accuracy: 0.9861\n",
      "Epoch 179/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0099 - accuracy: 0.9892 - val_loss: 0.0132 - val_accuracy: 0.9853\n",
      "Epoch 180/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0089 - accuracy: 0.9906 - val_loss: 0.0123 - val_accuracy: 0.9869\n",
      "Epoch 181/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0088 - accuracy: 0.9908 - val_loss: 0.0123 - val_accuracy: 0.9861\n",
      "Epoch 182/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0094 - accuracy: 0.9904 - val_loss: 0.0145 - val_accuracy: 0.9836\n",
      "Epoch 183/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0100 - accuracy: 0.9892 - val_loss: 0.0125 - val_accuracy: 0.9861\n",
      "Epoch 184/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0103 - accuracy: 0.9887 - val_loss: 0.0117 - val_accuracy: 0.9877\n",
      "Epoch 185/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0088 - accuracy: 0.9900 - val_loss: 0.0130 - val_accuracy: 0.9845\n",
      "Epoch 186/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0087 - accuracy: 0.9910 - val_loss: 0.0124 - val_accuracy: 0.9877\n",
      "Epoch 187/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0084 - accuracy: 0.9908 - val_loss: 0.0115 - val_accuracy: 0.9869\n",
      "Epoch 188/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0086 - accuracy: 0.9902 - val_loss: 0.0113 - val_accuracy: 0.9877\n",
      "Epoch 189/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0082 - accuracy: 0.9914 - val_loss: 0.0143 - val_accuracy: 0.9812\n",
      "Epoch 190/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0091 - accuracy: 0.9902 - val_loss: 0.0109 - val_accuracy: 0.9886\n",
      "Epoch 191/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0082 - accuracy: 0.9918 - val_loss: 0.0118 - val_accuracy: 0.9869\n",
      "Epoch 192/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0081 - accuracy: 0.9918 - val_loss: 0.0153 - val_accuracy: 0.9836\n",
      "Epoch 193/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0107 - accuracy: 0.9883 - val_loss: 0.0155 - val_accuracy: 0.9828\n",
      "Epoch 194/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0088 - accuracy: 0.9910 - val_loss: 0.0123 - val_accuracy: 0.9869\n",
      "Epoch 195/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0093 - accuracy: 0.9900 - val_loss: 0.0111 - val_accuracy: 0.9886\n",
      "Epoch 196/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0084 - accuracy: 0.9914 - val_loss: 0.0116 - val_accuracy: 0.9869\n",
      "Epoch 197/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0080 - accuracy: 0.9912 - val_loss: 0.0116 - val_accuracy: 0.9877\n",
      "Epoch 198/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0112 - accuracy: 0.9865 - val_loss: 0.0159 - val_accuracy: 0.9820\n",
      "Epoch 199/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0104 - accuracy: 0.9890 - val_loss: 0.0133 - val_accuracy: 0.9853\n",
      "Epoch 200/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0094 - accuracy: 0.9900 - val_loss: 0.0133 - val_accuracy: 0.9853\n",
      "Epoch 201/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0081 - accuracy: 0.9914 - val_loss: 0.0139 - val_accuracy: 0.9853\n",
      "Epoch 202/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0078 - accuracy: 0.9918 - val_loss: 0.0125 - val_accuracy: 0.9861\n",
      "Epoch 203/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0080 - accuracy: 0.9912 - val_loss: 0.0127 - val_accuracy: 0.9861\n",
      "Epoch 204/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0079 - accuracy: 0.9912 - val_loss: 0.0108 - val_accuracy: 0.9886\n",
      "Epoch 205/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0094 - accuracy: 0.9894 - val_loss: 0.0126 - val_accuracy: 0.9845\n",
      "Epoch 206/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0084 - accuracy: 0.9912 - val_loss: 0.0121 - val_accuracy: 0.9869\n",
      "Epoch 207/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0083 - accuracy: 0.9910 - val_loss: 0.0102 - val_accuracy: 0.9894\n",
      "Epoch 208/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0074 - accuracy: 0.9918 - val_loss: 0.0115 - val_accuracy: 0.9869\n",
      "Epoch 209/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0068 - accuracy: 0.9932 - val_loss: 0.0117 - val_accuracy: 0.9869\n",
      "Epoch 210/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0078 - accuracy: 0.9922 - val_loss: 0.0129 - val_accuracy: 0.9853\n",
      "Epoch 211/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0084 - accuracy: 0.9908 - val_loss: 0.0109 - val_accuracy: 0.9886\n",
      "Epoch 212/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0076 - accuracy: 0.9920 - val_loss: 0.0140 - val_accuracy: 0.9845\n",
      "Epoch 213/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0120 - accuracy: 0.9859 - val_loss: 0.0180 - val_accuracy: 0.9820\n",
      "Epoch 214/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0174 - accuracy: 0.9820 - val_loss: 0.0148 - val_accuracy: 0.9836\n",
      "Epoch 215/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0098 - accuracy: 0.9890 - val_loss: 0.0119 - val_accuracy: 0.9861\n",
      "Epoch 216/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0073 - accuracy: 0.9926 - val_loss: 0.0119 - val_accuracy: 0.9877\n",
      "Epoch 217/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0071 - accuracy: 0.9920 - val_loss: 0.0114 - val_accuracy: 0.9886\n",
      "Epoch 218/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0070 - accuracy: 0.9926 - val_loss: 0.0115 - val_accuracy: 0.9877\n",
      "Epoch 219/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0068 - accuracy: 0.9930 - val_loss: 0.0100 - val_accuracy: 0.9886\n",
      "Epoch 220/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0069 - accuracy: 0.9920 - val_loss: 0.0131 - val_accuracy: 0.9861\n",
      "Epoch 221/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0065 - accuracy: 0.9930 - val_loss: 0.0119 - val_accuracy: 0.9869\n",
      "Epoch 222/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0065 - accuracy: 0.9932 - val_loss: 0.0118 - val_accuracy: 0.9877\n",
      "Epoch 223/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0064 - accuracy: 0.9935 - val_loss: 0.0110 - val_accuracy: 0.9886\n",
      "Epoch 224/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0064 - accuracy: 0.9932 - val_loss: 0.0131 - val_accuracy: 0.9861\n",
      "Epoch 225/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0098 - accuracy: 0.9896 - val_loss: 0.0146 - val_accuracy: 0.9836\n",
      "Epoch 226/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0087 - accuracy: 0.9900 - val_loss: 0.0119 - val_accuracy: 0.9886\n",
      "Epoch 227/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0062 - accuracy: 0.9937 - val_loss: 0.0126 - val_accuracy: 0.9853\n",
      "Epoch 228/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0062 - accuracy: 0.9937 - val_loss: 0.0114 - val_accuracy: 0.9869\n",
      "Epoch 229/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0070 - accuracy: 0.9922 - val_loss: 0.0120 - val_accuracy: 0.9877\n",
      "Epoch 230/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0063 - accuracy: 0.9935 - val_loss: 0.0118 - val_accuracy: 0.9877\n",
      "Epoch 231/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0057 - accuracy: 0.9941 - val_loss: 0.0115 - val_accuracy: 0.9869\n",
      "Epoch 232/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0062 - accuracy: 0.9937 - val_loss: 0.0210 - val_accuracy: 0.9763\n",
      "Epoch 233/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0088 - accuracy: 0.9904 - val_loss: 0.0117 - val_accuracy: 0.9886\n",
      "Epoch 234/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0071 - accuracy: 0.9926 - val_loss: 0.0116 - val_accuracy: 0.9877\n",
      "Epoch 235/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0063 - accuracy: 0.9935 - val_loss: 0.0140 - val_accuracy: 0.9853\n",
      "Epoch 236/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0075 - accuracy: 0.9908 - val_loss: 0.0129 - val_accuracy: 0.9861\n",
      "Epoch 237/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0064 - accuracy: 0.9937 - val_loss: 0.0120 - val_accuracy: 0.9877\n",
      "Epoch 238/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0058 - accuracy: 0.9941 - val_loss: 0.0116 - val_accuracy: 0.9877\n",
      "Epoch 239/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0056 - accuracy: 0.9941 - val_loss: 0.0111 - val_accuracy: 0.9886\n",
      "Epoch 240/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0057 - accuracy: 0.9939 - val_loss: 0.0107 - val_accuracy: 0.9886\n",
      "Epoch 241/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0085 - accuracy: 0.9904 - val_loss: 0.0109 - val_accuracy: 0.9886\n",
      "Epoch 242/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0106 - accuracy: 0.9890 - val_loss: 0.0160 - val_accuracy: 0.9836\n",
      "Epoch 243/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0098 - accuracy: 0.9890 - val_loss: 0.0114 - val_accuracy: 0.9886\n",
      "Epoch 244/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0076 - accuracy: 0.9922 - val_loss: 0.0155 - val_accuracy: 0.9828\n",
      "Epoch 245/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0068 - accuracy: 0.9922 - val_loss: 0.0127 - val_accuracy: 0.9861\n",
      "Epoch 246/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0059 - accuracy: 0.9937 - val_loss: 0.0116 - val_accuracy: 0.9877\n",
      "Epoch 247/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0057 - accuracy: 0.9943 - val_loss: 0.0107 - val_accuracy: 0.9886\n",
      "Epoch 248/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0055 - accuracy: 0.9945 - val_loss: 0.0113 - val_accuracy: 0.9886\n",
      "Epoch 249/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0066 - accuracy: 0.9930 - val_loss: 0.0153 - val_accuracy: 0.9836\n",
      "Epoch 250/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0084 - accuracy: 0.9904 - val_loss: 0.0131 - val_accuracy: 0.9861\n",
      "Epoch 251/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0058 - accuracy: 0.9939 - val_loss: 0.0121 - val_accuracy: 0.9877\n",
      "Epoch 252/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0066 - accuracy: 0.9928 - val_loss: 0.0127 - val_accuracy: 0.9861\n",
      "Epoch 253/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0065 - accuracy: 0.9935 - val_loss: 0.0109 - val_accuracy: 0.9886\n",
      "Epoch 254/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0070 - accuracy: 0.9926 - val_loss: 0.0113 - val_accuracy: 0.9869\n",
      "Epoch 255/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0058 - accuracy: 0.9941 - val_loss: 0.0124 - val_accuracy: 0.9869\n",
      "Epoch 256/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0051 - accuracy: 0.9949 - val_loss: 0.0131 - val_accuracy: 0.9853\n",
      "Epoch 257/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0057 - accuracy: 0.9939 - val_loss: 0.0103 - val_accuracy: 0.9894\n",
      "Epoch 258/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0052 - accuracy: 0.9945 - val_loss: 0.0117 - val_accuracy: 0.9869\n",
      "Epoch 259/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0060 - accuracy: 0.9932 - val_loss: 0.0130 - val_accuracy: 0.9861\n",
      "Epoch 260/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0076 - accuracy: 0.9914 - val_loss: 0.0119 - val_accuracy: 0.9877\n",
      "Epoch 261/500\n",
      "153/153 [==============================] - 2s 16ms/step - loss: 0.0062 - accuracy: 0.9935 - val_loss: 0.0115 - val_accuracy: 0.9886\n",
      "Epoch 262/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0051 - accuracy: 0.9945 - val_loss: 0.0144 - val_accuracy: 0.9828\n",
      "Epoch 263/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0061 - accuracy: 0.9939 - val_loss: 0.0108 - val_accuracy: 0.9894\n",
      "Epoch 264/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0052 - accuracy: 0.9945 - val_loss: 0.0116 - val_accuracy: 0.9869\n",
      "Epoch 265/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0052 - accuracy: 0.9945 - val_loss: 0.0109 - val_accuracy: 0.9877\n",
      "Epoch 266/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0048 - accuracy: 0.9951 - val_loss: 0.0111 - val_accuracy: 0.9886\n",
      "Epoch 267/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0048 - accuracy: 0.9951 - val_loss: 0.0109 - val_accuracy: 0.9886\n",
      "Epoch 268/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0052 - accuracy: 0.9947 - val_loss: 0.0140 - val_accuracy: 0.9853\n",
      "Epoch 269/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0083 - accuracy: 0.9908 - val_loss: 0.0128 - val_accuracy: 0.9869\n",
      "Epoch 270/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0059 - accuracy: 0.9941 - val_loss: 0.0115 - val_accuracy: 0.9886\n",
      "Epoch 271/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0052 - accuracy: 0.9947 - val_loss: 0.0122 - val_accuracy: 0.9877\n",
      "Epoch 272/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0050 - accuracy: 0.9951 - val_loss: 0.0113 - val_accuracy: 0.9886\n",
      "Epoch 273/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0054 - accuracy: 0.9943 - val_loss: 0.0122 - val_accuracy: 0.9853\n",
      "Epoch 274/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0054 - accuracy: 0.9941 - val_loss: 0.0108 - val_accuracy: 0.9894\n",
      "Epoch 275/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0063 - accuracy: 0.9928 - val_loss: 0.0155 - val_accuracy: 0.9836\n",
      "Epoch 276/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0090 - accuracy: 0.9898 - val_loss: 0.0182 - val_accuracy: 0.9787\n",
      "Epoch 277/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0129 - accuracy: 0.9863 - val_loss: 0.0154 - val_accuracy: 0.9812\n",
      "Epoch 278/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0082 - accuracy: 0.9912 - val_loss: 0.0121 - val_accuracy: 0.9869\n",
      "Epoch 279/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0049 - accuracy: 0.9949 - val_loss: 0.0145 - val_accuracy: 0.9836\n",
      "Epoch 280/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0052 - accuracy: 0.9947 - val_loss: 0.0146 - val_accuracy: 0.9828\n",
      "Epoch 281/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0065 - accuracy: 0.9930 - val_loss: 0.0140 - val_accuracy: 0.9853\n",
      "Epoch 282/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0055 - accuracy: 0.9943 - val_loss: 0.0125 - val_accuracy: 0.9853\n",
      "Epoch 283/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0049 - accuracy: 0.9953 - val_loss: 0.0121 - val_accuracy: 0.9877\n",
      "Epoch 284/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0050 - accuracy: 0.9947 - val_loss: 0.0127 - val_accuracy: 0.9869\n",
      "Epoch 285/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0045 - accuracy: 0.9955 - val_loss: 0.0134 - val_accuracy: 0.9836\n",
      "Epoch 286/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0045 - accuracy: 0.9955 - val_loss: 0.0119 - val_accuracy: 0.9877\n",
      "Epoch 287/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0046 - accuracy: 0.9953 - val_loss: 0.0146 - val_accuracy: 0.9836\n",
      "Epoch 288/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0051 - accuracy: 0.9951 - val_loss: 0.0115 - val_accuracy: 0.9877\n",
      "Epoch 289/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0064 - accuracy: 0.9930 - val_loss: 0.0108 - val_accuracy: 0.9886\n",
      "Epoch 290/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0050 - accuracy: 0.9949 - val_loss: 0.0114 - val_accuracy: 0.9894\n",
      "Epoch 291/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0049 - accuracy: 0.9953 - val_loss: 0.0108 - val_accuracy: 0.9877\n",
      "Epoch 292/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0058 - accuracy: 0.9939 - val_loss: 0.0113 - val_accuracy: 0.9886\n",
      "Epoch 293/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0065 - accuracy: 0.9932 - val_loss: 0.0127 - val_accuracy: 0.9869\n",
      "Epoch 294/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0089 - accuracy: 0.9908 - val_loss: 0.0111 - val_accuracy: 0.9877\n",
      "Epoch 295/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0065 - accuracy: 0.9928 - val_loss: 0.0139 - val_accuracy: 0.9836\n",
      "Epoch 296/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0057 - accuracy: 0.9945 - val_loss: 0.0132 - val_accuracy: 0.9853\n",
      "Epoch 297/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0053 - accuracy: 0.9947 - val_loss: 0.0114 - val_accuracy: 0.9877\n",
      "Epoch 298/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0050 - accuracy: 0.9949 - val_loss: 0.0138 - val_accuracy: 0.9853\n",
      "Epoch 299/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0061 - accuracy: 0.9937 - val_loss: 0.0136 - val_accuracy: 0.9861\n",
      "Epoch 300/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0075 - accuracy: 0.9922 - val_loss: 0.0124 - val_accuracy: 0.9877\n",
      "Epoch 301/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0091 - accuracy: 0.9896 - val_loss: 0.0155 - val_accuracy: 0.9820\n",
      "Epoch 302/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0054 - accuracy: 0.9947 - val_loss: 0.0128 - val_accuracy: 0.9861\n",
      "Epoch 303/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0046 - accuracy: 0.9955 - val_loss: 0.0127 - val_accuracy: 0.9861\n",
      "Epoch 304/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0045 - accuracy: 0.9955 - val_loss: 0.0129 - val_accuracy: 0.9861\n",
      "Epoch 305/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0047 - accuracy: 0.9953 - val_loss: 0.0117 - val_accuracy: 0.9869\n",
      "Epoch 306/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0045 - accuracy: 0.9955 - val_loss: 0.0120 - val_accuracy: 0.9877\n",
      "Epoch 307/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0045 - accuracy: 0.9955 - val_loss: 0.0120 - val_accuracy: 0.9869\n",
      "Epoch 308/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0046 - accuracy: 0.9953 - val_loss: 0.0117 - val_accuracy: 0.9877\n",
      "Epoch 309/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0045 - accuracy: 0.9955 - val_loss: 0.0125 - val_accuracy: 0.9861\n",
      "Epoch 310/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0044 - accuracy: 0.9953 - val_loss: 0.0126 - val_accuracy: 0.9877\n",
      "Epoch 311/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0046 - accuracy: 0.9955 - val_loss: 0.0131 - val_accuracy: 0.9861\n",
      "Epoch 312/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0098 - accuracy: 0.9890 - val_loss: 0.0116 - val_accuracy: 0.9869\n",
      "Epoch 313/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0067 - accuracy: 0.9924 - val_loss: 0.0106 - val_accuracy: 0.9877\n",
      "Epoch 314/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0051 - accuracy: 0.9947 - val_loss: 0.0117 - val_accuracy: 0.9877\n",
      "Epoch 315/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0050 - accuracy: 0.9949 - val_loss: 0.0111 - val_accuracy: 0.9877\n",
      "Epoch 316/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0050 - accuracy: 0.9949 - val_loss: 0.0116 - val_accuracy: 0.9861\n",
      "Epoch 317/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0063 - accuracy: 0.9930 - val_loss: 0.0151 - val_accuracy: 0.9836\n",
      "Epoch 318/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0084 - accuracy: 0.9908 - val_loss: 0.0124 - val_accuracy: 0.9853\n",
      "Epoch 319/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0058 - accuracy: 0.9939 - val_loss: 0.0117 - val_accuracy: 0.9877\n",
      "Epoch 320/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0049 - accuracy: 0.9951 - val_loss: 0.0112 - val_accuracy: 0.9869\n",
      "Epoch 321/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0048 - accuracy: 0.9945 - val_loss: 0.0127 - val_accuracy: 0.9853\n",
      "Epoch 322/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0053 - accuracy: 0.9945 - val_loss: 0.0114 - val_accuracy: 0.9877\n",
      "Epoch 323/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9957 - val_loss: 0.0117 - val_accuracy: 0.9877\n",
      "Epoch 324/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0044 - accuracy: 0.9957 - val_loss: 0.0114 - val_accuracy: 0.9869\n",
      "Epoch 325/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9957 - val_loss: 0.0123 - val_accuracy: 0.9853\n",
      "Epoch 326/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9953 - val_loss: 0.0127 - val_accuracy: 0.9853\n",
      "Epoch 327/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0050 - accuracy: 0.9945 - val_loss: 0.0154 - val_accuracy: 0.9820\n",
      "Epoch 328/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0063 - accuracy: 0.9932 - val_loss: 0.0127 - val_accuracy: 0.9861\n",
      "Epoch 329/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0070 - accuracy: 0.9922 - val_loss: 0.0107 - val_accuracy: 0.9886\n",
      "Epoch 330/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0047 - accuracy: 0.9953 - val_loss: 0.0130 - val_accuracy: 0.9853\n",
      "Epoch 331/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0045 - accuracy: 0.9955 - val_loss: 0.0123 - val_accuracy: 0.9861\n",
      "Epoch 332/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9955 - val_loss: 0.0123 - val_accuracy: 0.9869\n",
      "Epoch 333/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9955 - val_loss: 0.0116 - val_accuracy: 0.9877\n",
      "Epoch 334/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9957 - val_loss: 0.0125 - val_accuracy: 0.9853\n",
      "Epoch 335/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0046 - accuracy: 0.9949 - val_loss: 0.0114 - val_accuracy: 0.9861\n",
      "Epoch 336/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0045 - accuracy: 0.9955 - val_loss: 0.0117 - val_accuracy: 0.9861\n",
      "Epoch 337/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9957 - val_loss: 0.0097 - val_accuracy: 0.9894\n",
      "Epoch 338/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0048 - accuracy: 0.9949 - val_loss: 0.0116 - val_accuracy: 0.9869\n",
      "Epoch 339/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9955 - val_loss: 0.0111 - val_accuracy: 0.9886\n",
      "Epoch 340/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0050 - accuracy: 0.9943 - val_loss: 0.0132 - val_accuracy: 0.9861\n",
      "Epoch 341/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0049 - accuracy: 0.9947 - val_loss: 0.0140 - val_accuracy: 0.9836\n",
      "Epoch 342/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0107 - accuracy: 0.9879 - val_loss: 0.0160 - val_accuracy: 0.9820\n",
      "Epoch 343/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0060 - accuracy: 0.9935 - val_loss: 0.0136 - val_accuracy: 0.9845\n",
      "Epoch 344/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0045 - accuracy: 0.9953 - val_loss: 0.0129 - val_accuracy: 0.9853\n",
      "Epoch 345/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9957 - val_loss: 0.0119 - val_accuracy: 0.9869\n",
      "Epoch 346/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0039 - accuracy: 0.9961 - val_loss: 0.0113 - val_accuracy: 0.9869\n",
      "Epoch 347/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0041 - accuracy: 0.9957 - val_loss: 0.0128 - val_accuracy: 0.9861\n",
      "Epoch 348/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0046 - accuracy: 0.9955 - val_loss: 0.0113 - val_accuracy: 0.9886\n",
      "Epoch 349/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0058 - accuracy: 0.9935 - val_loss: 0.0145 - val_accuracy: 0.9845\n",
      "Epoch 350/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9957 - val_loss: 0.0152 - val_accuracy: 0.9845\n",
      "Epoch 351/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0049 - accuracy: 0.9949 - val_loss: 0.0128 - val_accuracy: 0.9869\n",
      "Epoch 352/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0052 - accuracy: 0.9943 - val_loss: 0.0131 - val_accuracy: 0.9853\n",
      "Epoch 353/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0042 - accuracy: 0.9955 - val_loss: 0.0128 - val_accuracy: 0.9861\n",
      "Epoch 354/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0091 - accuracy: 0.9902 - val_loss: 0.0164 - val_accuracy: 0.9836\n",
      "Epoch 355/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0085 - accuracy: 0.9904 - val_loss: 0.0119 - val_accuracy: 0.9877\n",
      "Epoch 356/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0053 - accuracy: 0.9941 - val_loss: 0.0134 - val_accuracy: 0.9853\n",
      "Epoch 357/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0044 - accuracy: 0.9951 - val_loss: 0.0143 - val_accuracy: 0.9845\n",
      "Epoch 358/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0050 - accuracy: 0.9947 - val_loss: 0.0114 - val_accuracy: 0.9877\n",
      "Epoch 359/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0038 - accuracy: 0.9963 - val_loss: 0.0117 - val_accuracy: 0.9853\n",
      "Epoch 360/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0038 - accuracy: 0.9961 - val_loss: 0.0109 - val_accuracy: 0.9877\n",
      "Epoch 361/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0040 - accuracy: 0.9959 - val_loss: 0.0114 - val_accuracy: 0.9877\n",
      "Epoch 362/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9959 - val_loss: 0.0108 - val_accuracy: 0.9886\n",
      "Epoch 363/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0045 - accuracy: 0.9951 - val_loss: 0.0114 - val_accuracy: 0.9877\n",
      "Epoch 364/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9957 - val_loss: 0.0102 - val_accuracy: 0.9886\n",
      "Epoch 365/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0046 - accuracy: 0.9949 - val_loss: 0.0147 - val_accuracy: 0.9804\n",
      "Epoch 366/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0041 - accuracy: 0.9959 - val_loss: 0.0117 - val_accuracy: 0.9869\n",
      "Epoch 367/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0046 - accuracy: 0.9953 - val_loss: 0.0098 - val_accuracy: 0.9902\n",
      "Epoch 368/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0036 - accuracy: 0.9961 - val_loss: 0.0105 - val_accuracy: 0.9877\n",
      "Epoch 369/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0041 - accuracy: 0.9955 - val_loss: 0.0091 - val_accuracy: 0.9894\n",
      "Epoch 370/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0061 - accuracy: 0.9932 - val_loss: 0.0113 - val_accuracy: 0.9853\n",
      "Epoch 371/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0062 - accuracy: 0.9932 - val_loss: 0.0091 - val_accuracy: 0.9894\n",
      "Epoch 372/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0039 - accuracy: 0.9959 - val_loss: 0.0106 - val_accuracy: 0.9894\n",
      "Epoch 373/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0037 - accuracy: 0.9963 - val_loss: 0.0110 - val_accuracy: 0.9869\n",
      "Epoch 374/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0037 - accuracy: 0.9963 - val_loss: 0.0126 - val_accuracy: 0.9861\n",
      "Epoch 375/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0036 - accuracy: 0.9965 - val_loss: 0.0112 - val_accuracy: 0.9886\n",
      "Epoch 376/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0039 - accuracy: 0.9959 - val_loss: 0.0091 - val_accuracy: 0.9894\n",
      "Epoch 377/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0046 - accuracy: 0.9949 - val_loss: 0.0121 - val_accuracy: 0.9869\n",
      "Epoch 378/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0044 - accuracy: 0.9951 - val_loss: 0.0113 - val_accuracy: 0.9861\n",
      "Epoch 379/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0042 - accuracy: 0.9959 - val_loss: 0.0119 - val_accuracy: 0.9861\n",
      "Epoch 380/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0037 - accuracy: 0.9963 - val_loss: 0.0122 - val_accuracy: 0.9861\n",
      "Epoch 381/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0036 - accuracy: 0.9963 - val_loss: 0.0108 - val_accuracy: 0.9877\n",
      "Epoch 382/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0035 - accuracy: 0.9965 - val_loss: 0.0107 - val_accuracy: 0.9877\n",
      "Epoch 383/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0066 - accuracy: 0.9926 - val_loss: 0.0114 - val_accuracy: 0.9877\n",
      "Epoch 384/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0069 - accuracy: 0.9924 - val_loss: 0.0138 - val_accuracy: 0.9836\n",
      "Epoch 385/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0062 - accuracy: 0.9930 - val_loss: 0.0121 - val_accuracy: 0.9845\n",
      "Epoch 386/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9957 - val_loss: 0.0121 - val_accuracy: 0.9861\n",
      "Epoch 387/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0037 - accuracy: 0.9963 - val_loss: 0.0110 - val_accuracy: 0.9877\n",
      "Epoch 388/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0037 - accuracy: 0.9963 - val_loss: 0.0123 - val_accuracy: 0.9869\n",
      "Epoch 389/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0041 - accuracy: 0.9957 - val_loss: 0.0120 - val_accuracy: 0.9861\n",
      "Epoch 390/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0039 - accuracy: 0.9961 - val_loss: 0.0106 - val_accuracy: 0.9894\n",
      "Epoch 391/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0041 - accuracy: 0.9959 - val_loss: 0.0122 - val_accuracy: 0.9869\n",
      "Epoch 392/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9955 - val_loss: 0.0132 - val_accuracy: 0.9845\n",
      "Epoch 393/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0035 - accuracy: 0.9965 - val_loss: 0.0129 - val_accuracy: 0.9845\n",
      "Epoch 394/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0041 - accuracy: 0.9957 - val_loss: 0.0109 - val_accuracy: 0.9877\n",
      "Epoch 395/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0036 - accuracy: 0.9963 - val_loss: 0.0112 - val_accuracy: 0.9886\n",
      "Epoch 396/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0032 - accuracy: 0.9965 - val_loss: 0.0138 - val_accuracy: 0.9845\n",
      "Epoch 397/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0040 - accuracy: 0.9961 - val_loss: 0.0143 - val_accuracy: 0.9836\n",
      "Epoch 398/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0070 - accuracy: 0.9924 - val_loss: 0.0106 - val_accuracy: 0.9877\n",
      "Epoch 399/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0087 - accuracy: 0.9902 - val_loss: 0.0102 - val_accuracy: 0.9894\n",
      "Epoch 400/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0042 - accuracy: 0.9955 - val_loss: 0.0096 - val_accuracy: 0.9894\n",
      "Epoch 401/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0036 - accuracy: 0.9965 - val_loss: 0.0098 - val_accuracy: 0.9894\n",
      "Epoch 402/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0033 - accuracy: 0.9967 - val_loss: 0.0095 - val_accuracy: 0.9894\n",
      "Epoch 403/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0033 - accuracy: 0.9967 - val_loss: 0.0090 - val_accuracy: 0.9894\n",
      "Epoch 404/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0033 - accuracy: 0.9967 - val_loss: 0.0090 - val_accuracy: 0.9910\n",
      "Epoch 405/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0034 - accuracy: 0.9965 - val_loss: 0.0108 - val_accuracy: 0.9886\n",
      "Epoch 406/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9953 - val_loss: 0.0101 - val_accuracy: 0.9902\n",
      "Epoch 407/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0035 - accuracy: 0.9963 - val_loss: 0.0111 - val_accuracy: 0.9877\n",
      "Epoch 408/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0032 - accuracy: 0.9969 - val_loss: 0.0108 - val_accuracy: 0.9877\n",
      "Epoch 409/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0031 - accuracy: 0.9969 - val_loss: 0.0103 - val_accuracy: 0.9877\n",
      "Epoch 410/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0037 - accuracy: 0.9961 - val_loss: 0.0093 - val_accuracy: 0.9886\n",
      "Epoch 411/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0034 - accuracy: 0.9965 - val_loss: 0.0088 - val_accuracy: 0.9910\n",
      "Epoch 412/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0032 - accuracy: 0.9967 - val_loss: 0.0110 - val_accuracy: 0.9861\n",
      "Epoch 413/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0041 - accuracy: 0.9957 - val_loss: 0.0102 - val_accuracy: 0.9894\n",
      "Epoch 414/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0054 - accuracy: 0.9943 - val_loss: 0.0116 - val_accuracy: 0.9869\n",
      "Epoch 415/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0048 - accuracy: 0.9941 - val_loss: 0.0146 - val_accuracy: 0.9845\n",
      "Epoch 416/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0044 - accuracy: 0.9951 - val_loss: 0.0095 - val_accuracy: 0.9902\n",
      "Epoch 417/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0064 - accuracy: 0.9928 - val_loss: 0.0169 - val_accuracy: 0.9812\n",
      "Epoch 418/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0072 - accuracy: 0.9924 - val_loss: 0.0127 - val_accuracy: 0.9869\n",
      "Epoch 419/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0033 - accuracy: 0.9967 - val_loss: 0.0104 - val_accuracy: 0.9877\n",
      "Epoch 420/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0053 - accuracy: 0.9941 - val_loss: 0.0136 - val_accuracy: 0.9845\n",
      "Epoch 421/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0040 - accuracy: 0.9953 - val_loss: 0.0126 - val_accuracy: 0.9853\n",
      "Epoch 422/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0034 - accuracy: 0.9963 - val_loss: 0.0091 - val_accuracy: 0.9902\n",
      "Epoch 423/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0033 - accuracy: 0.9967 - val_loss: 0.0095 - val_accuracy: 0.9894\n",
      "Epoch 424/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0029 - accuracy: 0.9971 - val_loss: 0.0103 - val_accuracy: 0.9877\n",
      "Epoch 425/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0029 - accuracy: 0.9969 - val_loss: 0.0098 - val_accuracy: 0.9886\n",
      "Epoch 426/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0030 - accuracy: 0.9967 - val_loss: 0.0108 - val_accuracy: 0.9861\n",
      "Epoch 427/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0030 - accuracy: 0.9971 - val_loss: 0.0117 - val_accuracy: 0.9869\n",
      "Epoch 428/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0034 - accuracy: 0.9965 - val_loss: 0.0107 - val_accuracy: 0.9886\n",
      "Epoch 429/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0029 - accuracy: 0.9971 - val_loss: 0.0104 - val_accuracy: 0.9869\n",
      "Epoch 430/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0027 - accuracy: 0.9973 - val_loss: 0.0096 - val_accuracy: 0.9894\n",
      "Epoch 431/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0026 - accuracy: 0.9973 - val_loss: 0.0098 - val_accuracy: 0.9886\n",
      "Epoch 432/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0026 - accuracy: 0.9973 - val_loss: 0.0097 - val_accuracy: 0.9894\n",
      "Epoch 433/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0026 - accuracy: 0.9973 - val_loss: 0.0103 - val_accuracy: 0.9894\n",
      "Epoch 434/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0025 - accuracy: 0.9975 - val_loss: 0.0105 - val_accuracy: 0.9886\n",
      "Epoch 435/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0025 - accuracy: 0.9975 - val_loss: 0.0100 - val_accuracy: 0.9894\n",
      "Epoch 436/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0028 - accuracy: 0.9971 - val_loss: 0.0109 - val_accuracy: 0.9869\n",
      "Epoch 437/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0056 - accuracy: 0.9932 - val_loss: 0.0104 - val_accuracy: 0.9886\n",
      "Epoch 438/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0046 - accuracy: 0.9945 - val_loss: 0.0109 - val_accuracy: 0.9877\n",
      "Epoch 439/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0039 - accuracy: 0.9957 - val_loss: 0.0086 - val_accuracy: 0.9910\n",
      "Epoch 440/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0030 - accuracy: 0.9969 - val_loss: 0.0081 - val_accuracy: 0.9910\n",
      "Epoch 441/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0026 - accuracy: 0.9975 - val_loss: 0.0092 - val_accuracy: 0.9894\n",
      "Epoch 442/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0023 - accuracy: 0.9977 - val_loss: 0.0088 - val_accuracy: 0.9902\n",
      "Epoch 443/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0030 - accuracy: 0.9969 - val_loss: 0.0090 - val_accuracy: 0.9894\n",
      "Epoch 444/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0024 - accuracy: 0.9975 - val_loss: 0.0085 - val_accuracy: 0.9910\n",
      "Epoch 445/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0034 - accuracy: 0.9965 - val_loss: 0.0138 - val_accuracy: 0.9853\n",
      "Epoch 446/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0050 - accuracy: 0.9943 - val_loss: 0.0142 - val_accuracy: 0.9836\n",
      "Epoch 447/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0035 - accuracy: 0.9967 - val_loss: 0.0120 - val_accuracy: 0.9869\n",
      "Epoch 448/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0030 - accuracy: 0.9967 - val_loss: 0.0122 - val_accuracy: 0.9869\n",
      "Epoch 449/500\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0027 - accuracy: 0.9971 - val_loss: 0.0095 - val_accuracy: 0.9902\n",
      "Epoch 450/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0025 - accuracy: 0.9975 - val_loss: 0.0100 - val_accuracy: 0.9894\n",
      "Epoch 451/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0024 - accuracy: 0.9975 - val_loss: 0.0105 - val_accuracy: 0.9894\n",
      "Epoch 452/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0023 - accuracy: 0.9977 - val_loss: 0.0101 - val_accuracy: 0.9894\n",
      "Epoch 453/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0023 - accuracy: 0.9977 - val_loss: 0.0107 - val_accuracy: 0.9886\n",
      "Epoch 454/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0023 - accuracy: 0.9977 - val_loss: 0.0107 - val_accuracy: 0.9894\n",
      "Epoch 455/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0024 - accuracy: 0.9975 - val_loss: 0.0105 - val_accuracy: 0.9886\n",
      "Epoch 456/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0024 - accuracy: 0.9975 - val_loss: 0.0104 - val_accuracy: 0.9886\n",
      "Epoch 457/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0023 - accuracy: 0.9977 - val_loss: 0.0097 - val_accuracy: 0.9894\n",
      "Epoch 458/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0024 - accuracy: 0.9975 - val_loss: 0.0104 - val_accuracy: 0.9886\n",
      "Epoch 459/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0023 - accuracy: 0.9977 - val_loss: 0.0101 - val_accuracy: 0.9877\n",
      "Epoch 460/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0025 - accuracy: 0.9975 - val_loss: 0.0110 - val_accuracy: 0.9869\n",
      "Epoch 461/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0062 - accuracy: 0.9928 - val_loss: 0.0167 - val_accuracy: 0.9812\n",
      "Epoch 462/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0059 - accuracy: 0.9937 - val_loss: 0.0125 - val_accuracy: 0.9861\n",
      "Epoch 463/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0026 - accuracy: 0.9971 - val_loss: 0.0128 - val_accuracy: 0.9853\n",
      "Epoch 464/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9953 - val_loss: 0.0095 - val_accuracy: 0.9894\n",
      "Epoch 465/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0035 - accuracy: 0.9961 - val_loss: 0.0107 - val_accuracy: 0.9886\n",
      "Epoch 466/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0043 - accuracy: 0.9953 - val_loss: 0.0097 - val_accuracy: 0.9894\n",
      "Epoch 467/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0025 - accuracy: 0.9975 - val_loss: 0.0108 - val_accuracy: 0.9869\n",
      "Epoch 468/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0029 - accuracy: 0.9971 - val_loss: 0.0094 - val_accuracy: 0.9886\n",
      "Epoch 469/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0025 - accuracy: 0.9975 - val_loss: 0.0115 - val_accuracy: 0.9861\n",
      "Epoch 470/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0029 - accuracy: 0.9969 - val_loss: 0.0118 - val_accuracy: 0.9853\n",
      "Epoch 471/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0022 - accuracy: 0.9980 - val_loss: 0.0121 - val_accuracy: 0.9853\n",
      "Epoch 472/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0032 - accuracy: 0.9965 - val_loss: 0.0122 - val_accuracy: 0.9853\n",
      "Epoch 473/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0030 - accuracy: 0.9969 - val_loss: 0.0133 - val_accuracy: 0.9853\n",
      "Epoch 474/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0028 - accuracy: 0.9971 - val_loss: 0.0117 - val_accuracy: 0.9869\n",
      "Epoch 475/500\n",
      "153/153 [==============================] - 2s 13ms/step - loss: 0.0019 - accuracy: 0.9982 - val_loss: 0.0120 - val_accuracy: 0.9869\n",
      "Epoch 476/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0027 - accuracy: 0.9969 - val_loss: 0.0106 - val_accuracy: 0.9869\n",
      "Epoch 477/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0036 - accuracy: 0.9959 - val_loss: 0.0140 - val_accuracy: 0.9845\n",
      "Epoch 478/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0034 - accuracy: 0.9965 - val_loss: 0.0135 - val_accuracy: 0.9845\n",
      "Epoch 479/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0024 - accuracy: 0.9973 - val_loss: 0.0134 - val_accuracy: 0.9861\n",
      "Epoch 480/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0020 - accuracy: 0.9980 - val_loss: 0.0133 - val_accuracy: 0.9836\n",
      "Epoch 481/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0019 - accuracy: 0.9982 - val_loss: 0.0127 - val_accuracy: 0.9853\n",
      "Epoch 482/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0019 - accuracy: 0.9982 - val_loss: 0.0121 - val_accuracy: 0.9853\n",
      "Epoch 483/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0020 - accuracy: 0.9982 - val_loss: 0.0131 - val_accuracy: 0.9845\n",
      "Epoch 484/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0022 - accuracy: 0.9977 - val_loss: 0.0117 - val_accuracy: 0.9861\n",
      "Epoch 485/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0024 - accuracy: 0.9977 - val_loss: 0.0120 - val_accuracy: 0.9877\n",
      "Epoch 486/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0032 - accuracy: 0.9967 - val_loss: 0.0101 - val_accuracy: 0.9886\n",
      "Epoch 487/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0021 - accuracy: 0.9977 - val_loss: 0.0116 - val_accuracy: 0.9869\n",
      "Epoch 488/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0019 - accuracy: 0.9982 - val_loss: 0.0106 - val_accuracy: 0.9886\n",
      "Epoch 489/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0019 - accuracy: 0.9982 - val_loss: 0.0124 - val_accuracy: 0.9853\n",
      "Epoch 490/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0023 - accuracy: 0.9975 - val_loss: 0.0102 - val_accuracy: 0.9886\n",
      "Epoch 491/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0032 - accuracy: 0.9965 - val_loss: 0.0126 - val_accuracy: 0.9869\n",
      "Epoch 492/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0078 - accuracy: 0.9912 - val_loss: 0.0157 - val_accuracy: 0.9828\n",
      "Epoch 493/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0094 - accuracy: 0.9902 - val_loss: 0.0121 - val_accuracy: 0.9861\n",
      "Epoch 494/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0049 - accuracy: 0.9945 - val_loss: 0.0122 - val_accuracy: 0.9861\n",
      "Epoch 495/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0026 - accuracy: 0.9971 - val_loss: 0.0117 - val_accuracy: 0.9845\n",
      "Epoch 496/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0023 - accuracy: 0.9977 - val_loss: 0.0110 - val_accuracy: 0.9877\n",
      "Epoch 497/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0021 - accuracy: 0.9980 - val_loss: 0.0118 - val_accuracy: 0.9869\n",
      "Epoch 498/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0019 - accuracy: 0.9982 - val_loss: 0.0120 - val_accuracy: 0.9853\n",
      "Epoch 499/500\n",
      "153/153 [==============================] - 2s 15ms/step - loss: 0.0019 - accuracy: 0.9982 - val_loss: 0.0115 - val_accuracy: 0.9869\n",
      "Epoch 500/500\n",
      "153/153 [==============================] - 2s 14ms/step - loss: 0.0018 - accuracy: 0.9982 - val_loss: 0.0115 - val_accuracy: 0.9861\n"
     ]
    }
   ],
   "source": [
    "epochs = 500\n",
    "batch_size = 32\n",
    "\n",
    "callbacks = [\n",
    "    keras.callbacks.ModelCheckpoint(\n",
    "        \"best_model_500_epoches_feature_selection.h5\", save_best_only=True, monitor=\"val_loss\"\n",
    "    ),\n",
    "    keras.callbacks.ReduceLROnPlateau(\n",
    "        monitor=\"val_loss\", factor=0.5, patience=20, min_lr=0.0001\n",
    "    ),\n",
    "    keras.callbacks.EarlyStopping(monitor=\"val_loss\", patience=50, verbose=1),\n",
    "]\n",
    "model.compile(\n",
    "    optimizer=\"adam\",\n",
    "    loss = 'mean_squared_error', \n",
    "    metrics = [\"accuracy\"],\n",
    ")\n",
    "history1 = model.fit(\n",
    "    X_train,\n",
    "    y_train,\n",
    "    batch_size=batch_size,\n",
    "    epochs=epochs,\n",
    "#     callbacks=callbacks,\n",
    "    validation_split=0.2,\n",
    "    verbose=1,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "fb9a38a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_train_validation_graph(history, file_name):\n",
    "    loss = history.history['loss']\n",
    "    val_loss = history.history['val_loss']\n",
    "    \n",
    "    acc = history.history['accuracy']\n",
    "    val_acc = history.history['val_accuracy']\n",
    "\n",
    "    epochs_range = range(len(loss))  \n",
    "    \n",
    "\n",
    "    plt.figure(figsize=(8, 6), dpi=100)\n",
    "    \n",
    "    plt.subplot(2, 1, 1)\n",
    "\n",
    "    plt.plot(epochs_range, acc, label='Training Accuracy')\n",
    "    plt.plot(epochs_range, val_acc, label='Validation Accuracy')\n",
    "    plt.xlabel('Epoch')\n",
    "    plt.ylabel('Accuracy')\n",
    "    plt.legend(loc='lower right')\n",
    "    plt.grid(True)\n",
    "   \n",
    "    plt.title('Training Vs Validation Accuracy')\n",
    "    \n",
    "    plt.subplot(2, 1, 2)      \n",
    "    plt.plot(epochs_range, loss, label='Training Loss')\n",
    "    plt.plot(epochs_range, val_loss, label='Validation Loss')\n",
    "    plt.legend(loc='upper right')\n",
    "    plt.xlabel('Epoch')\n",
    "    plt.ylabel('MSE')\n",
    "    plt.title('Training Vs Validation Loss')   \n",
    "    plt.tight_layout(pad=1.0)\n",
    "    plt.grid(True)\n",
    "    plt.legend()\n",
    "    \n",
    "    plt.savefig(file_name + \".png\")\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "546e0563",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_train_validation_graph(history1, \"Training_Vs_Validation_graph_lstm_feature_selection_500_epochs\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "8ab227e0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "34/34 [==============================] - 0s 7ms/step - loss: 0.0121 - accuracy: 0.9870\n",
      "Test accuracy: 0.9870\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model\n",
    "loss, accuracy = model.evaluate(X_test, y_test)\n",
    "print(f'Test accuracy: {accuracy:.4f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "00f97424",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "34/34 [==============================] - 0s 6ms/step\n"
     ]
    }
   ],
   "source": [
    "y_prediction = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "a6f2e6e4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1079, 10, 10)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "b1eec9ec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((1079, 1), (1079,))"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_prediction.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "8efce12b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1079,)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_predict = y_prediction.flatten()\n",
    "y_predict.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "06a83ac1",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_predict = list(np.where(np.array(y_predict) > 0.5, 1, 0 ))\n",
    "y_predict = np.array(y_predict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "0ae9a525",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((1079,), (1079,))"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_predict.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cd790287",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "644e88f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataframe = pd.DataFrame({\n",
    "    \"Actual_Load\":y_test,\n",
    "    \"Predicted_Load\":y_predict\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "230ddbff",
   "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>Actual_Load</th>\n",
       "      <th>Predicted_Load</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1074</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1075</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1076</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1077</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1078</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1079 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Actual_Load  Predicted_Load\n",
       "0               1               1\n",
       "1               0               0\n",
       "2               0               0\n",
       "3               0               0\n",
       "4               0               0\n",
       "...           ...             ...\n",
       "1074            1               0\n",
       "1075            1               0\n",
       "1076            0               0\n",
       "1077            0               0\n",
       "1078            0               0\n",
       "\n",
       "[1079 rows x 2 columns]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "6e2bc2dc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Actual_Load\n",
       "0    1056\n",
       "1      23\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataframe[\"Actual_Load\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "295bf402",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Predicted_Load\n",
       "0    1068\n",
       "1      11\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataframe[\"Predicted_Load\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "7ef77aaa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Squared Error: 0.012974976830398516\n",
      "Root Mean Squared Error: 0.11390775579563718\n",
      "Mean Absolute Error: 0.012974976830398516\n",
      "R-squared Score: 0.3780467720685111\n"
     ]
    }
   ],
   "source": [
    "# Calculate evaluation metrics\n",
    "mse = mean_squared_error(y_test, y_predict)\n",
    "rmse = np.sqrt(mse)\n",
    "mae = mean_absolute_error(y_test, y_predict)\n",
    "r2 = r2_score(y_test, y_predict)\n",
    "\n",
    "print(\"Mean Squared Error:\", mse)\n",
    "print(\"Root Mean Squared Error:\", rmse)\n",
    "print(\"Mean Absolute Error:\", mae)\n",
    "print(\"R-squared Score:\", r2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef8fbd9e",
   "metadata": {},
   "source": [
    "# Confustion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "7fb30a3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def display_evaluation(yt, yp):\n",
    "    TN, FP, FN, TP = confusion_matrix(yt, yp).ravel()\n",
    "\n",
    "    print(\"Model: \", model)\n",
    "\n",
    "    print(\"Confusion Matix Values:\")\n",
    "    print(\"TN: \", TN)\n",
    "    print(\"FP: \", FP)\n",
    "    print(\"FN: \", FN)\n",
    "    print(\"TP: \", TP)\n",
    "\n",
    "    SE = TP/(TP+FN)  #sensitivity, recall, hit rate, or true positive rate (TPR)\n",
    "    SP = TN/(TN+FP)  #specificity, selectivity or true negative rate (TNR)\n",
    "    AC = (TP+TN)/(TP+TN+FP+FN)   #accuracy\n",
    "    PR = TP/(TP+FP)   #precision, positive predictive value (PPV)\n",
    "    BCR = 1/2*(SE+SP)   #balanced classification rate\n",
    "    BER = 1-BCR   #balanced error rate\n",
    "    F1 = 2*PR*SE/(PR+SE)   #F1 score\n",
    "    FNR = 1-SE   #miss rate or false negative rate (FNR)\n",
    "    FPR = 1-SP   #fall-out or false positive rate (FPR)\n",
    "    TS = TP/(TP+FP+FN)   #Threat score (TS) or Critical Success Index (CSI)\n",
    "    DC = 2*TP/(2*TP+FP+FN) #Dice Coefficient\n",
    "\n",
    "    print(\"\\nEvaluation Parameters:\")\n",
    "    print('Accuracy:{:.4f}'.format(AC))\n",
    "    print('Sensitivity:{:.4f}'.format(SE))\n",
    "    print('Specificity: {:.4f}'.format(SP))\n",
    "    print('Precision: {:.4f}'.format(PR))\n",
    "    print('F1 Score: {:.4f}'.format(F1))\n",
    "    print('BCR: {:.4f}'.format(BCR))\n",
    "    print('BER: {:.4f}'.format(BER))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "751d8c26",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model:  <keras.engine.sequential.Sequential object at 0x7f048ce8dd30>\n",
      "Confusion Matix Values:\n",
      "TN:  1055\n",
      "FP:  1\n",
      "FN:  13\n",
      "TP:  10\n",
      "\n",
      "Evaluation Parameters:\n",
      "Accuracy:0.9870\n",
      "Sensitivity:0.4348\n",
      "Specificity: 0.9991\n",
      "Precision: 0.9091\n",
      "F1 Score: 0.5882\n",
      "BCR: 0.7169\n",
      "BER: 0.2831\n"
     ]
    }
   ],
   "source": [
    "display_evaluation(y_test, y_predict)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7235c7ca",
   "metadata": {},
   "source": [
    "# ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "313c8901",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAskAAAIsCAYAAAAAvDTcAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAACSbUlEQVR4nOzdeVhUZcMG8HvYNwFBVNwBXyD3HVFzQ0XcUNwBwSXBtDRNS7PMLbcySzB3xQU1TcCdNE1RydLUNM0NTAURkH2Rdc73hx8nR0BhGDgD3L/req+3eeacMzczojcPzzlHJgiCACIiIiIiEmlIHYCIiIiISN2wJBMRERERvYYlmYiIiIjoNSzJRERERESvYUkmIiIiInoNSzIRERER0WtYkomIiIiIXsOSTERERET0GpZkIiIiIqLXsCQTVVO9e/fG3LlzpY5R7YwbNw7jxo2TOsZb+fn5wc7ODomJiVJHUTt2dnbw8/NTybGioqJgZ2eHoKAglRwPAG7cuIEWLVogOjpaZcdUtZkzZ2LGjBlSxyB6Iy2pAxBVRUFBQZg3b574WFNTE+bm5ujatStmzpyJOnXqSJhOvWVmZmL79u0IDQ3F48ePoaWlBTs7O4waNQqurq6QyWRSR3yrBw8e4MSJExg2bBgaNGggdRwF+fn5CAkJQUhICO7evYvMzEzUrl0bDg4OcHd3R8uWLaWOWGZHjhxBQkICxo8fL3UUUUVmWrNmDQYOHIj69euLY+PGjcMff/whPtbV1UXjxo0xfPhweHl5QUOj8JxZUlISNm3ahDNnzuDp06fQ19dHy5Yt4enpiV69ehX52unp6QgICMDJkyfx5MkT5Ofno1GjRujRowe8vLzEv/smT56M4cOH486dO7C3t1fxO0CkGizJROVo+vTpaNCgAXJycnD9+nUEBwfjzz//xNGjR6GrqytpttDQULUrnM+fP8f48eMRERGBAQMGwNPTE9nZ2Th58iQ+/fRTnDt3Dt988w00NTWljvpGDx48gL+/Pzp16lSoJG/dulWiVEBWVhY++OADnD9/Hh07doSvry9MTEwQHR2NEydOIDg4GGfPnkXdunUly6gKR48exf3798utkN64caPUfwaLy1S/fn3cuHEDWlqq+ef4n3/+QXh4OPbt21foubp162LWrFkAXhbgo0ePYvny5UhKSsLMmTMVto2MjMT48eORmJgINzc3tGzZEqmpqThy5AimTJmCiRMn4tNPP1XY58mTJxg/fjxiYmLQv39/jB49Gtra2rh79y5++ukn/PLLL/j5558BAM2aNUOLFi2wbds2rFq1SiVfO5GqsSQTlaPu3buLM3MjR45EzZo1sXnzZpw+fRoDBgyQNJuOjk6Fv2Z2dja0tbWLnLUCgE8//RQRERHw9/eHk5OTOO7l5YWVK1di27ZteOedd+Dj41NRkQG8nN02MDBQybGkeN8LrFq1CufPn8e8efMKlbUPPvgAAQEBFZpHEARkZ2dDT0+vQl9XGXK5HLm5udDV1VXpD7gymUylxzt48CDq1auHNm3aFHquRo0acHV1FR+PHTsWLi4u2LVrF6ZPny4W/9zcXMyYMQOpqakIDAxE69atxX3Gjx+P2bNnY9u2bWjZsqX491heXh4++OADJCQkYOfOnejQoYPCa8+cORObN29WGHNxcYGfnx8yMjJgaGioqreASGW4JpmoAhX8w/HkyROF8YiICEyfPh2dOnVCy5Yt4ebmhtOnTxfaPzU1FcuWLUPv3r3RokULdO/eHZ988onCutGcnBysXbsWffv2RYsWLdCjRw+sWrUKOTk5Csd6dU3yzZs3YWdnh+Dg4EKvef78edjZ2eHXX38Vx2JjYzFv3jx06dIFLVq0wMCBA/HTTz8p7Pf777/Dzs4Ox44dw5o1a/Duu++idevWSE9PL/K9uX79Oi5cuIBhw4YpFOQCH3/8MZo0aYItW7YgKysLwH/rObdu3YqAgAD06tULrVq1gqenJ+7du1foGCV5n4OCgmBnZ4c//vgDCxcuhKOjI3r06AEAiI6OxsKFC+Hs7IxWrVrBwcEB06dPR1RUlML+BWstvby8YGdnBzs7O/z+++8ACq9JLnifjh8/jvXr14s/WHl7e+PRo0eFvobAwEA4OTmhVatWGDFiBK5cuVKidc7Pnj3Djz/+iK5duxY5w6qpqYlJkyYVmkVOS0vD3Llz0aFDB7Rv3x7z5s3DixcvFLY5ePAgvLy84OjoiBYtWmDAgAHYs2dPodfo3bs3fH19cf78ebi5uaFVq1bijGdJjwEA586dg6enJ9q2bYt27dph+PDhOHLkiPj+nj17FtHR0eJ737t3b3Hfkn5/2NnZYfHixTh8+DAGDhyIli1b4vz58+Jzr65JTk9Px1dffSV+Xzo6OmLChAm4devWWzMVtyY5IiICM2bMQOfOndGqVSs4OztjzZo1Rb4frzp9+jQ6d+5cot8S6erqokWLFsjIyEBCQoI4fvLkSdy7dw+TJ09WKMjAyz8nixcvhrGxscJ7cPLkSdy5cwdTpkwpVJABwMjIqNBsdZcuXZCZmYnw8PC3ZiWSAmeSiSpQwYk0xsbG4tj9+/cxduxY1KlTB5MnT4aBgQFOnDiBadOmwc/PD3379gUAZGRkwMPDAxERERg+fDiaNWuGpKQknDlzBrGxsTAzM4NcLsf777+PP//8E6NGjYKNjQ3u3buHHTt24N9//8UPP/xQZK6WLVuiYcOG4jraVx0/fhwmJibo1q0bgJdLIkaNGgWZTAYPDw+YmZkhLCwM8+fPR3p6eqEC9sMPP0BbWxuTJk1CTk4OtLW1i8xQUMKHDh1a5PNaWloYNGgQ/P39cfXqVXTp0kV8LiQkBBkZGXB3d0d2djZ27doFb29vHDlyBLVq1SrV+1xg0aJFMDMzw7Rp05CZmQng5Q8T165dw8CBA1G3bl1ER0dj79698PLywrFjx6Cvr4+OHTti3Lhx2LVrF6ZMmQJra2sAgI2NTZFfV4HNmzdDJpNh4sSJSE9Px5YtWzB79mwcOHBA3GbPnj1YvHgxOnTogPHjxyM6OhrTpk2DsbHxW5dIhIWFIS8vD0OGDHnjdq/76KOP0KBBA8yaNQu3b9/GgQMHYGZmhjlz5ojb7N27F//73//Qu3dvaGlp4ddff8WiRYsgCAI8PDwUjvfw4UN8/PHHGD16NEaNGgUrK6tSHSMoKAifffYZ/ve//8HX1xc1atTAP//8g/Pnz2Pw4MGYMmUK0tLS8OzZM/G8gIJZytJ+f1y6dAknTpyAh4cHatasqbDG91Vffvklfv75Z3h6esLGxgbJycn4888/ERERgebNm78xU1Hu3LkDDw8PaGlpYfTo0ahfvz4eP36MM2fOFCqar4qNjcXTp0/RrFmzYrd5XXR0NGQymcLfSWfOnAFQ/PdijRo14OTkhODgYDx69AiNGzcWf9h8dab6bZo2bQo9PT1cvXq10PcfkVoQiEjlDh48KNja2grh4eFCQkKCEBMTI4SGhgqdO3cWWrRoIcTExIjbent7C4MGDRKys7PFMblcLowePVro16+fOPb9998Ltra2wsmTJwu9nlwuFwRBEEJCQgR7e3vh8uXLCs/v3btXsLW1Ff78809xrFevXsKnn34qPl69erXQvHlzITk5WRzLzs4WOnToIMybN08c++yzz4SuXbsKiYmJCq8xc+ZMoX379sKLFy8EQRCES5cuCba2toKTk5M49iZTp04VbG1thZSUlGK3OXnypGBrayvs3LlTEARBePLkiWBrayu0atVKePbsmbjdX3/9Jdja2grLli0Tx0r6Phd8dmPHjhXy8vIUXr+or+PatWuCra2tEBwcLI6dOHFCsLW1FS5dulRoe09PT8HT01N8XPA+ubi4KGTbsWOHYGtrK9y9e1cQhJefRadOnYThw4cLubm54nZBQUGCra2twjGLsmzZMsHW1la4ffv2G7crsHbtWsHW1lbhsxcEQZg2bZrQqVMnhbGi3peJEycKTk5OCmO9evUSbG1thbCwsELbl+QYqampQtu2bYWRI0cKWVlZCtsWfA8IgiD4+PgIvXr1KnS80nx/2NraCvb29sL9+/cLHcfW1lZYu3at+Lh9+/bCokWLCm33quIyFfwZPnjwoDjm4eEhtG3bVoiOji72ayxKeHi4YGtrK5w5c6bQc56enkL//v2FhIQEISEhQYiIiBBWrlwp2NraCj4+Pgrburq6Cu3bt3/ja23fvl2wtbUVTp8+LQiCIAwdOvSt+xSlX79+wnvvvVfq/YgqApdbEJWj8ePHi7+unz59OvT19bF+/Xpx1i85ORmXLl2Ci4sL0tPTkZiYiMTERCQlJaFbt274999/ERsbC+DlrzPt7e2LnHEp+NVqaGgobGxsYG1tLR4rMTERnTt3BgDxV/5FGTBgAHJzc3Hy5Elx7OLFi0hNTRXXHQqCgJMnT6J3794QBEHhNbp164a0tDTxV8wFhg4dWqI1pxkZGQDePMNW8NzrSzb69OmjcMWQVq1aoXXr1jh37hyA0r3PBUaNGlXo5KxXv47c3FwkJSWhUaNGMDY2xu3bt9/6Nb6Jm5ubwnrl15fm/P3330hOTsaoUaMUTvIaPHgwTExM3nr8gvestGs/x4wZo/C4Q4cOSE5OVvgMXn1f0tLSkJiYiE6dOuHJkydIS0tT2L9BgwZ49913C71OSY5x8eJFZGRkwMfHp9A63pIsLyjt90fHjh3RtGnTtx7X2NgYf/31V6E/Q8pITEzE5cuXMXz4cNSrV0/hubd9jUlJSWKeokRGRsLR0RGOjo5wcXHB1q1b0bt3byxfvlxhu5KsEX79ezE9PV2pdcUmJiZibiJ1w+UWROVowYIFsLKyQlpaGg4ePIjLly8rFKHHjx9DEAR8//33+P7774s8RkJCAurUqYPHjx+jX79+b3y9R48eISIiAo6OjsUeqzj29vawtrbGiRMnMHLkSAAvl1rUrFlTLBGJiYlITU3Fjz/+iB9//LHI47x+Xd2SXgKt4B/YjIyMYv+RL65IN27cuNC2TZo0wYkTJwCU7n1+U+6srCxs3LgRQUFBiI2NhSAI4nOvl8HSer0QFbwHqampAICnT58CABo1aqSwnZaWVrHLAF5lZGQE4L/3sKy5UlJSxGP++eef8PPzw/Xr1wutV05LS0ONGjXEx8X9eSjJMR4/fgwA+N///leqr6FAab8/Svpnd/bs2Zg7dy569uyJ5s2bo0ePHhg6dCgaNmxY6owFPxTZ2tqWet8Cr/65fFX9+vWxdOlSyOVyPH78GBs2bEBSUlKhHzgMDQ3fWlxf/140MjIqdK5FSbOq21V2iAqwJBOVo1atWolXt+jTpw/c3d3x8ccfIzQ0FIaGhpDL5QCAiRMnFjm7BhQuRW8il8tha2urcI3mV71t3eqAAQOwYcMGJCYmwsjICGfOnMHAgQPFmcuCvEOGDCm0drmAnZ2dwuOSXrnAxsYGv/zyC+7evYuOHTsWuc3du3cBoESze69S5n0u6ooDS5YsQVBQELy9vdGmTRvUqFEDMpkMM2fOLLaYlFRxV/wo63ELFKyNvnv3Lt555x2V5Xr8+DHGjx8Pa2trzJ07F5aWltDW1sa5c+cQEBAgvvcFivrzUNpjKKu03x8l/bM7YMAAdOjQAadOncLFixexdetWbN68GX5+fuJJnxWhZs2aAP77wep1BgYGCmv527VrBzc3N6xZswaff/65OG5jY4N//vkHT58+LfRDUoHXvxetra1x+/ZtxMTEwNLSssSZU1NTi/whl0gdsCQTVRBNTU3MmjULXl5eCAwMhI+PjzjTpK2trfCPV1EaNWqE+/fvv3WbO3fuwNHRUanZmQEDBsDf3x8nT55ErVq1kJ6ejoEDB4rPm5mZieX+bXlLq2fPnti4cSNCQkKKLMn5+fk4cuQITExM0K5dO4XniroKxL///ivOsJbmfX6Tn3/+GUOHDlW4U2F2dnahWeTymBkrKCuPHz8WZ/aBl5feKrhqwpt0794dmpqaOHLkSLEnZCnjzJkzyMnJwfr16xUK1ZuW9ih7jIIfZO7fv//GYlXc+1/W7483qV27Njw8PODh4YGEhAQMGzYMGzZsEEtySV+v4M9qUVdneZuCH4RevdrKm9jb22PIkCHYt28fJk6cKL73PXv2xNGjRxESEoKpU6cW2i89PR2nT5+GtbW1+Dn06tULR48exeHDh+Hr61ui18/Ly0NMTIzC1UeI1AnXJBNVIAcHB7Rq1Qo7duxAdnY2zM3N0alTJ/z444+Ii4srtP2rSxf69euHO3fu4NSpU4W2K5jVc3FxQWxsLPbv319om6ysLPEqDcWxsbGBra0tjh8/juPHj8PCwkKhsGpqasLZ2Rk///xzkf+Il+UWxu3atUOXLl0QFBSkcLm5AmvWrMG///6L9957r9AM3y+//KKwHvTGjRv466+/0L17dwAo1fv8JkXdQGLXrl3Iz89XGNPX1wdQ9iUYr2rRogVMTU2xf/9+5OXlieNHjhxBSkrKW/e3tLTEyJEjceHCBezatavQ83K5HNu2bcOzZ89KlavgPXl96cnBgwdVfoxu3brB0NAQGzduRHZ2tsJzr+6rr69f5Htf1u+PouTn5xd6LXNzc9SuXVvhsnLFZXqdmZkZOnbsiIMHD4pLbAq87bcKderUgaWlJf7+++8S53/vvfeQl5eH7du3i2POzs5o2rQpNm/ejJs3bypsL5fL8eWXXyIlJQUffPCBwj62trbYsGEDrl27Vuh10tPTC13C7sGDB8jOzkbbtm1LnJeoInEmmaiCTZo0CTNmzEBQUBDGjh2LL7/8Eu7u7hg8eDBGjRqFhg0b4vnz57h+/TqePXuGw4cPi/v9/PPPmDFjBoYPH47mzZsjJSUFZ86cwaJFi2Bvbw9XV1ecOHECX375JX7//Xe0a9cO+fn5iIyMRGhoKLZs2fLW2w4PGDAAa9euha6uLkaMGFHo1+0ff/wxfv/9d4waNQojR45E06ZNkZKSglu3buG3335TuPVtaa1cuRLjx4/H1KlTMWjQIHTo0AE5OTk4efIk/vjjDwwYMACTJk0qtF+jRo0wduxYjB07Fjk5Odi5cydMTU3x3nvviduU9H1+k549e+LQoUMwMjJC06ZNcf36dYSHh8PU1FRhu3feeQeamprYvHkz0tLSoKOjg86dO8Pc3Fzp90ZHRwcffvghlixZAm9vb7i4uCA6OhpBQUElXpIzd+5cPHnyBEuXLsXJkyfRq1cvGBsbIyYmBqGhoYiMjFT4zUFJdO3aFdra2pgyZQrGjBmDjIwMHDhwAObm5oiPj1fpMYyMjDBv3jx8/vnnGDFiBAYNGgRjY2PcuXMHWVlZWLlyJQCgefPmOH78OJYvX46WLVvCwMAAvXv3Vsn3x+syMjLQo0cPODs7w97eHgYGBggPD8fNmzcVfuNQXKaifP755xg7diyGDRuG0aNHo0GDBoiOjsbZs2dx6NChN+ZxcnLCqVOnSrzWt2nTpujRowd++uknTJ06FTVr1oSOjg7Wrl0Lb29vuLu7w83NDS1atEBaWhqOHj2KW7duYeLEiQp/VrS1teHv748JEybA09MT/fv3R7t27aCtrY379+/j6NGjMDY2VriEXXh4OPT19VX+WykiVWFJJqpg/fr1Q6NGjbBt2zaMGjUKTZs2xcGDB+Hv74/g4GAkJyfDzMwMzZo1w7Rp08T9DA0NERgYCD8/P5w6dQrBwcEwNzeHo6OjeMKZhoYG1q1bh4CAABw6dAinTp2Cvr4+GjRogHHjxonXpH2TAQMG4LvvvsOLFy/g4uJS6PlatWrhwIEDWLduHU6dOoW9e/fC1NQUTZs2xezZs8v03tSuXRsHDhzA9u3bERoaipMnT0JTUxN2dnZYsWIFhg4dWuQ//EOHDoWGhgZ27NiBhIQEtGrVCl988QVq164tblPS9/lN5s+fDw0NDRw5cgTZ2dlo164dtm/frlDGAcDCwgKLFi3Cxo0bMX/+fOTn52Pnzp1lKskA4OnpCUEQsH37dqxcuRL29vZYv349li5dWqK7tunr62Pz5s0ICgpCSEgIfvjhB2RlZaF27dpwcHDAN998o3DyYklYW1tj7dq1+O6777By5UrUqlULY8eOhZmZGT777DOVH2PkyJEwNzfHpk2b8MMPP0BLSwvW1tYK1+d2d3fHP//8g6CgIAQEBKB+/fro3bu3Sr4/Xqenp4exY8fi4sWLOHnyJARBQKNGjcQfyt6WqSj29vbYv38/vv/+e+zduxfZ2dmoV69ekd+Prxs+fDh2796NP//8s8ibehRl0qRJOHv2LHbv3o0PP/wQwMvfKh0+fBibNm3CmTNnEBQUBD09PbRo0QLr168vMnvjxo0REhKCgIAAnDp1CqdPn4ZcLkfjxo0xcuTIQje8CQ0NRd++fcUTQInUjUxQ1VkhREQVLCoqCk5OTvjkk0+KnGGuDuRyORwdHdG3b18sXbpU6jikBry9vVG7dm18/fXXUkcp1j///INhw4YhODi4VCeSElUkrkkmIqoksrOzC61LDQkJQXJyMjp16iRRKlI3s2bNwokTJ8Q7fKqjTZs2wdnZmQWZ1BqXWxARVRLXr1/H8uXL0b9/f5iamuL27dv46aefYGtri/79+0sdj9RE69atS3XynhReP4mPSB2xJBMRVRL169dH3bp1sWvXLqSkpMDExASurq6YPXu2wk1qiIio7LgmmYiIiIjoNVyTTERERET0GpZkIiIiIqLXcE1yMQpuYmBhYSF1FCIiIiIqQnx8PHR0dHDlyhWVH5sluRjZ2dmFbjVLREREROojLy/vrbdsVxZLcjEK7tR1+vRpiZMQERERUVGcnJzK7dhck0xERERE9BqWZCIiIiKi17AkExERERG9hiWZiIiIiOg1LMlERERERK9hSSYiIiIieg1LMhERERHRa1iSiYiIiIhew5JMRERERPQalmQiIiIiotewJBMRERERvYYlmYiIiIjoNSzJRERERESvYUkmIiIiInqNWpXkR48eYcGCBXB1dUWzZs0waNCgEu0nCAI2bdqEnj17olWrVhg9ejSuX79evmGJiIiIqMpSq5J8//59nDt3Do0bN4aNjU2J99u8eTPWrl2L8ePHY+PGjbCwsMDEiRPx5MmTckxLRERERFWVWpXk3r1749y5c1i7di2aN29eon2ys7OxceNGTJw4EePHj4ejoyO+/fZbmJqaYuvWreWcmIiIiIiqIrUqyRoapY9z9epVpKenw8XFRRzT0dFB3759ERYWpsp4RERERFRNqFVJVkZkZCQAwNraWmHcxsYGT58+RVZWlhSxiIiIiKgS05I6QFmlpqZCR0cHurq6CuPGxsYQBAEpKSnQ09OTKB0RERGpxN0DQPgCICdN6iSkBgQB2HLBDilxz2BSu265vEalL8lERERUDYQvABLvSJ2C1MDzDANMPjAYIX+/AyurQzApp9ep9CXZ2NgYOTk5yM7OVphNTk1NhUwmg4lJeb11REREVGEKZpBlGoChpbRZSDKnbteH944eiEkxBPByRrm8VPqSXLAW+eHDh7C3txfHIyMjUa9ePS61ICIiqkoMLQHfKKlTUAXLysrDZ5+dxpq1l8Qxc3N91K1rVG6vWelP3GvXrh2MjIxw4sQJcSw3NxcnT55E9+7dJUxGRERERGV161YcHBy2YM2a/wpyv342uHnzfRgYaJfb66rVTPKLFy9w7tw5AEB0dDTS09MRGhoKAOjUqRPMzMzg7e2Np0+f4tSpUwAAXV1d+Pr6ws/PD2ZmZrC1tcXevXuRnJyMSZMmSfa1EBEREVHZXLjwGH377kJWVh4AQFdXEytX9sGHHzpAQ0NWrq+tViU5ISEBM2bMUBgreLxz5044ODhALpcjPz9fYZvJkydDEARs27YNiYmJeOedd7B161Y0bNiwwrITERERkWp17FgPdnbm+OuvWLRoURt79rihZcs6FfLaMkEozyXPlZeTkxMA4PTp0xInISIiImxsAKRHA0b1uSa5mrl1Kw7btl3DV185QU9PcX63PPtapV+TTERERESVX2ZmLqZPP4Fbt+IUxps3r43Vq50LFeTyplbLLYiIiIio+rl2LQbu7kG4c+c5zp17hD/+eA+6utLWVM4kExEREZEk5HIBX399EQ4OW3DnznMAwP37CfjzzxiJk3EmmYiIiIgkEBWVCm/vEJw581Aca9fOEoGBbrC3ryVhspdYkomIiIioQh04cAu+vkeRlJQFAJDJgE8+6YrFi3tBR0dT4nQvsSQTERERUYVIS8vG9OmhCAi4Lo41aGCMnTuHolcvK+mCFYElmYiIiIgqxK1b8di58y/x8ciRzbBx4yDUrKkvYaqi8cQ9IiIiIqoQnTs3wOefvwsjIx0EBLjixx9HqGVBBjiTTERERETlJCoqFZaWRtDU/G9e9osvemDChLZo0sRUumAlwJlkIiIiIlIpQRCwe/cNNGu2Dt98E67wnJaWhtoXZIAlmYiIiIhUKDk5C+7uQRg3LhhpaTn4/PNfcfWq9Nc9Li0utyAiqq7uHgDCFwA5aVInIXq7jMpXsqqjsLBHGDcuGI8fp4hj7u4t0bSpmYSplMOSTERUXYUvABLvSJ2CqHR0akidgIqQm5uPhQvPYvnyCxCEl2MmJrrYuHEQRo9uIW04JbEkExFVVwUzyDINwNBS2ixEJaFTA+i6ROoU9Jp79xLg4RGEK1eeimM9ejTGzp3D0KiRiYTJyoYlmYioujO0BHyjpE5BRJXQ+fOP0L9/IDIzcwG8PClvyZJemDOni8IVLSojlmQiIiIiUkq7dpZo2NAYd+8mwNbWHIGBbujQoZ7UsVSCJZmIiIiIlGJoqIM9e4Zj8+Y/8c03/WBoqCN1JJWp3PPgRERERFQhsrLy8Omnp3D/foLCeLt2lli/flCVKsgAZ5KJiIiI6C1u3YqDu3sQbtyIxdmzj3DhwgRoa2tKHatccSaZiIiIiIokCAL8/H5Hhw6bceNGLADgr7+eKVzJoqriTDIRERERFfLsWTomTjyEEyceiGMtWtTGnj1uaNmyjoTJKgZLMhEREREpOHr0HiZOPIT4+ExxbMYMB6xY0Qd6etWjPlaPr5KIiIiI3iozMxezZ5/E+vVXxLG6dY0QEOAKZ+emEiareCzJRERERAQAuHLlqUJBHjLEDlu2DIaFhaGEqaTBE/eIiIiICADQvXtjfPyxI/T1tbBhw0CEhIyulgUZ4EwyERERUbUVG5uO2rUNIZPJxLGvvuoNH5/2sLU1lzCZ9DiTTERERFQNHThwC++8sw7+/n8ojOvqalX7ggywJBMRERFVK2lp2Zgw4RBGjfoJSUlZmDPnFP7+O07qWGqHyy2IiIiIqolLl6Lg4RGEyMgkcWzIEDvUr19DwlTqiSWZiIiIqIrLy5Nj2bLzWLz4HPLzBQCAkZEO/P1d4OXVWmFNMr3EkkxERERUhT18mARPz2CEhz8Rxzp3boDdu4fBxsZMwmTqjSWZSJ3dPQCELwBy0qROQlVRRozUCYionF248BgDBgQiLS0HAKChIcMXX3TH5593h5YWT017E5ZkInUWvgBIvCN1CqrqdLgWkaiqatWqDmrVMkBaWg6srEyxe7cbunRpKHWsSoElmUidFcwgyzQAQ0tps1DVpFMD6LpE6hREVE6MjXWxe7cbNm++iu+/7w9jY12pI1UaLMlElYGhJeAbJXUKIiJSY7m5+Vi27DwmTGiLRo1MxPEuXRpy9lgJLMlEREREldy9ewnw8AjClStP8euv/+L0aS9oanLNcVnw3SMiIiKqpARBwObNf6Jt2424cuUpAODixSe4fPmpxMkqP84kExEREVVCz59nYvLkIwgJ+e8Eb1tbc+zZ44b27etJmKxqYEkmIiIiqmROnYqAt3cIYmLSxTFf3/ZYvbofDA11JExWdbAkExEREVUSWVl5+Oyz01iz5pI4VquWAbZsGQxXV3sJk1U9LMlERERElcSFC48VCrKzsw22b3eFpSWvd65qPHGPiIiIqJLo08caU6a0h66uJr7/vj+OH/dgQS4nnEkmIiIiUlNJSS9gaqoHmUwmjq1e7Yxp0zqhRYvaEiar+jiTTERERKSGjh69Bzs7f2zffl1h3MBAmwW5ArAkExEREamRzMxcTJ16DIMH70V8fCamTz+BBw8SpY5V7XC5BREREZGauHYtBu7uQbhz57k45uRkDRMTXQlTVU8syUREREQSk8sFrF4djvnzzyA3Vw4A0NfXwpo1zvDxaa+wJpkqBksyERERkYSiolLh7R2CM2ceimPt2lkiMNAN9va1JExWvbEkExEREUnk4sXHGDx4L5KSsgAAMhnwySddsXhxL+joaEqcrnpjSSYiIiKSiL19LejrayMpKQsNGhhj165h6NmzidSxCCzJVNncPQCELwBy0qROUjEyYqROQERE5cjc3AA7dw7F5s1XsX79QNSsqS91JPp/LMlUuYQvABLvSJ2i4unwbkpERJVdXp4cq1eHw9u7DerWNRLHnZys4eRkLWEyKgpLMlUuBTPIMg3A0FLaLBVFpwbQdYnUKYiIqAwePkyCp2cwwsOf4OzZRzh+3J1XrFBzLMlUORlaAr5RUqcgIiJ6I0EQEBh4E1OnHkNaWg4A4OTJCPzxRzQcHBpInI7ehCWZiIiIqBwkJ2fh/fePYd++v8UxKytT7N7txoJcCbAkExEREalYWNgjjBsXjMePU8QxL6/W8PNzgbEx755XGbAkExEREalIbm4+Fi48i+XLL0AQXo6ZmOhi48ZBGD26hbThqFRYkomIiIhU5NSpSCxbdkF83KNHY+zcOQyNGplImIqUoSF1ACIiIqKqYsCA/8HLqzW0tDSwfLkTTp/2YkGupDiTTERERKSktLRs1KihuMbYz88FM2Y4oF27anKp0iqKM8lERERESjh1KgJ2dv4KV68AAGNjXRbkKoAlmYiIiKgUsrLyMGvWz+jXbzdiYtIxZcpRhatYUNXA5RZEREREJXTrVhzc3YNw40asOObg0AA6OpoSpqLywJJMRERE9BaCIMDf/w/MmXMK2dn5AABdXU2sXNkHH37oAA0N3mK6qmFJJiIiInqDZ8/SMXHiIZw48UAca9GiNvbscUPLlnUkTEbliSWZiIiIqBjh4U8wdOg+xMdnimMzZjhgxYo+0NNjjarK+OkSERERFcPauqb433XrGiEgwBXOzk0lTEQVhVe3ICIiIipG3bpG2LbNFUOG2OHGjSksyNUISzIRERERALlcgJ/f70hIyFQYHzTIFiEho2FhYShRMpICl1uQ6t09AIQvAHLSVH/sjBjVH5OIiKq9qKhUeHuH4MyZh/j1139x8OAoyGT/XbHi1f+m6oElmVQvfAGQeKd8X0OnRvken4iIqo2ffroNH58jSErKAgCEhNzBlStP0bFjfYmTkZRYkkn1CmaQZRqAYTncllOnBtB1ieqPS0RE1UpaWjZmzAjF9u3XxbEGDYyxa9cwFmRiSaZyZGgJ+EZJnYKIiKiQS5ei4OkZhIiIJHFs1Kjm2LBhIGrW1JcwGakLlmQiIiKqNvLy5Fi+/DwWLTqH/HwBAGBkpIN16wZg3LhWXHtMIpZkIiIiqjaOHbuHBQvOio87d26A3buHwcbGTLpQpJbU7hJwERERmDBhAtq0aYOuXbti1apVyMnJeet+SUlJWLBgAXr27Ik2bdpg0KBB2Lt3bwUkJiIiospiyBA7jBjRDBoaMixc2APnz09gQaYiqdVMckpKCry9vdGkSRP4+fkhNjYWK1asQFZWFhYsWPDGfWfMmIHIyEjMmjULlpaWCAsLw8KFC6GpqYlRo0ZV0FdARERE6iQrK0/h9tEymQwbNw7CrFmd4ejYUMJkpO7UqiTv27cPGRkZ8Pf3h6mpKQAgPz8fixYtgq+vL+rUqVPkfvHx8fj999+xfPlyuLm5AQAcHR1x8+ZNHDt2jCWZiIioGgoLewQvr2CsXeuCIUPsxHEzM30WZHortVpuERYWBkdHR7EgA4CLiwvkcjkuXrxY7H55eXkAgBo1FK+da2RkBEEQyiUrERERqafc3HzMn38aPXsG4NGjFEyadBjPnqVLHYsqGbUqyZGRkbC2tlYYMzY2hoWFBSIjI4vdz9LSEt26dcOGDRvw4MEDpKen4/jx47h48SI8PDzKOzYRERGpiXv3EtClyzYsW3YBBfNkzZtbID9fLm0wqnTUarlFamoqjI2NC42bmJggJSXljfv6+flh5syZGDhwIABAU1MTn3/+OZydncslKxEREakPQRCwZctVfPTRz8jMzAUAaGlpYMmSXpgzpws0NdVqXpAqAbUqycoSBAHz5s3Dv//+i9WrV8PCwgLh4eFYtmwZTExMxOJMREREVc/z55mYPPkIQkLuiGO2tubYs8cN7dvXkzAZVWZqVZKNjY2RlpZWaDwlJQUmJibF7nf27FmEhobi8OHDsLN7uTDfwcEBCQkJWLFiBUsyERFRFXXpUhTc3H5ETMx/a459fdtj9ep+MDTUkTAZVXZq9bsHa2vrQmuP09LSEB8fX2it8qsePHgATU1N2NraKoy/8847iIuLw4sXL8olLxEREUmrfv0aePHi5Qn85ub6CAkZjQ0bBrEgU5mpVUnu3r07wsPDkZqaKo6FhoZCQ0MDXbt2LXa/+vXrIz8/H3fv3lUYv3XrFszNzaGvz3uwExERVUUNG5pg06ZB6NfPBjdvvg9XV3upI1EVoVYlecyYMTA0NMS0adNw4cIFHDx4EKtWrcKYMWMUrpHs7e2Nvn37io+7d++OevXqYfr06Th06BB+++03fP311wgODoanp6cUXwoRERGpWMHJeSkpWQrjI0c2R2ioBywtaxSzJ1HpqdWaZBMTE+zYsQNLlizBtGnTYGhoiBEjRmDmzJkK28nlcuTn54uPjYyMEBAQgDVr1uCbb75BWloaGjRogLlz57IkExERVQHPnqVj4sRDOHHiAcLCHmHnzmEKz8tkMomSUVUlE3i3jSI5OTkBAE6fPi1xkkpoYwMgPRowqg/4RkmdhoiIKrmjR+9h4sRDiI/PFMf+/NMH7dpZSpiK1EF59jW1mkmmUrp7AAhfAOQUviKIpDJipE5ARERVQGZmLmbPPon166+IY3XqGCIgYCgLMpU7luTKLHwBkHjn7dtJRYdrw4iISDnXrsXA3T0Id+48F8cGD7bF1q1DYGFhKGEyqi5YkiuzV2eQjepLl6MoOjWArkukTkFERJWMXC5g9epwzJ9/Brm5L28lra+vhTVrnOHj055rj6nCsCRXBVz7S0REVURQ0D/45JNfxMdt29bFnj3DYW9fS8JUVB2p1SXgiIiIqHobPvwdDBjwP8hkwCefdMGlS++xIJMkOJNMREREksnNzYe2tqb4WCaTYdu2Ibh9Ox69ellJmIyqO84kExERkSQuXYpCs2Y/4NSpCIXxOnWMWJBJcizJREREVKHy8uRYsuQcunXbhgcPEuHtHYKEhMy370hUgbjcgoiIiCrMw4dJGDcuGBcvPhHHGjc2RUZGLszNJQxG9BqWZCIiIip3giAgMPAmpk49hrS0HACAhoYMCxZ0x/z53aGlxV9uk3phSSYiIqJylZychfffP4Z9+/4Wx6ysTLF7txu6dGkoYTKi4rEkExERUbm5fDkaI0YcwOPHKeKYl1dr+Pm5wNhYV8JkRG/GkkxERETlxsxMH4mJLwAApqZ62LBhIEaPbiFxKqK34wIgIiIiKjc2NmZYu7Y/evRojBs3prAgU6XBkkxEREQq8fLkvBvIyMhRGB8/vg3OnPFGw4YmEiUjKj2WZCIiIiqz588z4ea2H56ewfj445MKz8lkMmhoyCRKRqQclmQiIiIqk5MnI9Cq1XqEhNwBAGzc+CeuX38mcSqismFJJiIiIqVkZeVh1qyf4ey8GzEx6QAAc3N9hISMRps2dSVOR1Q2vLoFERERldqtW3Fwdw/CjRux4li/fjYICHCFpWUNCZMRqQZLckW4ewAIXwDkpKn2uBkxqj0eERHRWwiCAH//PzBnzilkZ+cDAHR1NbFyZR98+KED1x5TlcGSXBHCFwCJd8rv+Dr8iZ2IiCrG/v23MH16qPi4RYva2LPHDS1b1pEwFZHqsSRXhIIZZJkGYGip2mPr1AC6LlHtMYmIiIoxYkQz9Oz5J86e/RczZjhgxYo+0NNjnaCqh3+qK5KhJeAbJXUKIiKiEsvPl0NT87/z/DU1NbBz51DcuhWP/v2bSpiMqHzx6hZERERUpGvXYtCmzUZcvPhYYbxhQxMWZKryWJKJiIhIgVwu4OuvL8LBYQv+/jsOnp7BSEnJkjoWUYXicgsiIiISRUWlwts7BGfOPBTHzMz0kZSUBRMTPQmTEVUslmQiIiICABw4cAu+vkeRlPRy1lgmAz75pCsWL+4FHR1NidMRVSyWZCIiomouLS0bM2aEYvv26+JYgwbG2LVrGHr2bCJZLiIpsSQTERFVY1evxmDUqAOIiEgSx0aObIaNGwehZk19CZMRSYslmYiIqBrT09NCdPTL6/kbGenA398FXl6tIZPxznlUvfHqFkRERNVYs2YW+OabvujcuQGuX/eFt3cbFmQisCQTERFVG4IgIDj4H2Rl5SmMT53aEefPT4CNjZlEyYjUD0syERFRNZCcnAV39yC4ue3H/PmnFZ6TyWTQ0mIlIHoVvyOIiIiquLCwR2jdegP27fsbAPDtt5fw999xEqciUm8syURERFVUbm4+5s8/jZ49A/D4cQoAwMREF/v2DUeLFrUlTkek3nh1CyIioiro3r0EeHgE4cqVp+JYjx6NsXPnMDRqZCJhMqLKgSWZiIioChEEAVu3XsOMGaHIzMwFAGhpaWDJkl6YM6cLNDX5S2SikmBJJiIiqkL27v0bkycfER/b2ppjzx43tG9fT8JURJUPS3Jp3D0AhC8ActJKt19GTPnkISIies2oUc3h5/cHLl2Kgo9PO3z7rTMMDXWkjkVU6bAkl0b4AiDxjvL769RQXRYiIiK8XF7x6s0/tLQ0sHv3MNy8GYehQ+0lTEZUubEkl0bBDLJMAzC0LN2+OjWArktUn4mIiKqtW7fiMH78IWzYMFBhOYWNjRlvDEJURmUqyTk5Obh16xYSEhLQrl07mJlVk29IQ0vAN0rqFEREVE0JggB//z8wZ84pZGfnw909CFev+nBZBZEKKX2K686dO9GtWze4u7vjww8/xN27dwEAiYmJcHBwwE8//aSykERERPTSs2fpGDBgD6ZPD0V2dj4AQEdHE3FxGRInI6palCrJBw8exLJly/Duu+/iq6++giAI4nNmZmbo3Lkzjh8/rrKQREREBBw9eg+tWq1HaOgDcWzGDAdcvjwZVlY1JUxGVPUotdxi+/btcHJywurVq5GUlFTo+ebNm2PXrl1lDkdERERAZmYuZs8+ifXrr4hjdesaISDAFc7OTSVMRlR1KVWSHz16hHHjxhX7vKmpKZKTk5XNRERERP/vxo1YjB79E+7ceS6ODRlihy1bBsPCwlDCZERVm1Il2djYuMgZ5AIPHjyAhYWF0qGIiIjopbw8OSIiEgEA+vpaWLPGGT4+7RUu+0ZEqqfUmuTu3btj//79SE1NLfTc/fv3ceDAAfTu3bvM4YiIiKq7du0ssXRpb7RrZ4mrV33h69uBBZmoAihVkj/66CPk5+dj0KBB+O677yCTyRASEoLZs2dj+PDhMDMzw9SpU1WdlYiIqMr7+ecHyM3NVxibPbsLfvttEuzta0mUiqj6Uaok16lTB0FBQXj33Xdx4sQJCIKAQ4cO4ddff8XAgQOxf//+6nPNZCIiIhVIS8vGxImH0L9/IBYvPqfwnIaGDDo6mhIlI6qelL6ZiLm5Ob766it89dVXSExMhFwuh5mZGTQ0lL70MhERUbV06VIUPD2DEBHx8nyfZcsuYOzYlmjWjOf3EElFqUY7b948/PXXX+JjMzMz1KpVSyzIN27cwLx581STkIiIqIrKy5Nj8eJz6NZtm1iQjYx0sH27K955h0sriKSkVEkODg7G48ePi30+KioKISEhymYiIiKq8h4+TEKPHgH48suzyM9/eVOuzp0b4Pp1X3h5tebJeUQSU3q5xZvExcVBT0+vPA5NRERUqQmCgMDAm5g69RjS0nIAvFxzvGBBd8yf3x1aWly2SKQOSlySf/nlF5w+fVp8vH//foSHhxfaLi0tDeHh4WjRooVqEhIREVUhe/bcxLhxweJjKytT7N7thi5dGkqYioheV+KSHBERgdDQUACATCbDX3/9hb///lthG5lMBgMDA3Ts2BFz585VbVIiIqIqYMSIZvjmm99w/fozeHm1hp+fC4yNdaWORUSvKXFJ9vX1ha+vLwDA3t4eX331FQYPHlxuwYiIiKoCQRAU1hfr6mohMNANN2/GYvRo/taVSF0ptSb5zp07qs5BRERU5dy7l4D33juM9esHonnz2uJ4s2YWvLwbkZrj2QFEREQqJggCtmy5irZtN+L8+cdwdw9CVlae1LGIqBSUvrrFuXPnEBAQgNu3byMtLQ2CIBTa5p9//ilTOCIiosomISETkycfQXDwf791zcrKw9OnabC2rilhMiIqDaVmkn/++WdMmTIFz58/x4ABAyCXyzFw4EAMGDAAenp6sLOzw7Rp01SdlYiISK2dOhWBli3XKxRkH592uHrVhwWZqJJRaiZ548aNaNWqFfbs2YOUlBTs3bsXw4cPh6OjI6KiojB69Gg0aNBA1VmJiIjUUlZWHj777DTWrLkkjpmb62Pr1iFwdbWXMBkRKUupmeSIiAgMGDAAmpqa0NJ62bPz8l6utWrQoAHGjh2LzZs3qy4lERGRmrp9Ox4ODlsUCnK/fja4efN9FmSiSkypmWQ9PT1oa2sDAIyNjaGjo4P4+Hjx+Vq1aiEqKko1CYmIiNRYSkoWbt2KAwDo6mpi5co++PBDB2ho8LbSRJWZUjPJVlZWiIiIEB+/8847OHToEPLy8pCdnY2jR4/C0tJSZSGJiIjUlaNjQ3z+eXe0aFEbly9PxowZnVmQiaoApUpy3759cfr0aeTkvLzn/JQpU/DHH3+gY8eO6Ny5M65cuQIfHx+VBiUiIlIHYWGPkJ8vVxj7/PPuuHx5Mlq2rCNRKiJSNaWWW0yaNAmTJk0SH/fq1Qu7du3CyZMnoampiR49eqBz584qC0lERCS1zMxczJ59EuvXX8Hy5U6YO7eb+JyWlga0tHjrAaKqROnrJL+uQ4cO6NChg/g4PT0dRkZGqjo8ERGRZK5di4G7exDu3HkOAPjii18xbJg97OxqSZyMiMqLyn/sTUhIwLfffotevXqp+tBEREQVSi4X8M034XBw2CIWZH19Lfj7u8DW1lzidERUnko1k5yQkICQkBA8fvwYJiYm6NevH1q0aAEAiI2Nxfr16xEcHIzs7Gx06tSpXAITERFVhKioVHh7h+DMmYfiWLt2lggMdIO9PWeQiaq6EpfkiIgIeHp6Ijk5WbwF9ZYtW/D1119DJpNh/vz5yMnJQb9+/TBp0iSxPBMREVU2P/10Gz4+R5CUlAUAkMmATz7pisWLe0FHR1PidERUEUpckr///ntkZmbiyy+/RIcOHRAVFYXly5dj2bJlSEtLQ69evTB79mw0bNiwPPMSERGVq927b2DcuGDxcYMGxti1axh69mwiXSgiqnAlLslXrlzB2LFjMWbMGABA06ZNoampicmTJ2PYsGFYvnx5uYUkIiKqKG5u7+Crr87jzp3nGDmyGTZuHISaNfWljkVEFazEJTk5ORl2dnYKY/b2L2+32adPH9WmIiIikoiBgTb27HHDjRux8PJqDZmMNwYhqo5KfHULuVwOLS3FTl3w2MDAQLWpiIiIKsDDh0lwdt6N+/cTFMbbtrWEt3cbFmSiaqxUV7f4+++/oaurKz7OyMiATCbDn3/+ibS0tELb9+vXr9SBIiIisHTpUly7dg2GhoZwdXXFRx99BB0dnbfuGxsbi2+//Rbnzp1DZmYm6tevj/fffx9DhgwpdQ4iIqq6BEFAYOBNTJ16DGlpOfDwCMLFixOhrc2T8ojopVKV5B07dmDHjh2Fxv39/QuNyWQy/PPPP6UKk5KSAm9vbzRp0gR+fn6IjY3FihUrkJWVhQULFrxx37i4OIwePRpWVlZYsmQJjIyMcP/+ffHW2WV29wCQHq2aYxERkWSSk7Pw/vvHsG/f3+LY8+eZePIkFdbWNSVMRkTqpMQleefOneWZAwCwb98+ZGRkwN/fH6ampgCA/Px8LFq0CL6+vqhTp06x+3799deoW7cutmzZAk3NlzMBjo6OqgsX/kpJ16mhuuMSEVGFCQt7hHHjgvH4cYo45uXVGn5+LjA21n3DnkRU3ZS4JFfEzUHCwsLg6OgoFmQAcHFxwZdffomLFy/Czc2tyP3S09Nx4sQJLFu2TCzIKpfzynKSrkvK5zWIiKhc5ObmY+HCs1i+/AL+/1L/MDHRxcaNgzB6NK/rT0SFqfy21GURGRkJa2trhTFjY2NYWFggMjKy2P1u3bqF3NxcaGlpwdPTE82bN0fXrl3x9ddfIzc3V7UhjeoDtiNUe0wiIio39+4loEuXbVi27L+C3KNHY9y48T4LMhEVS61KcmpqKoyNjQuNm5iYICUlpYg9Xnr+/DkA4PPPP0eLFi2wdetWeHt7Y8eOHVi7dm255SUiIvX35EkKrlx5CgDQ0tLA8uVOOH3aC40amUicjIjUWalO3FNXcrkcANClSxfMnTsXANC5c2dkZGRg27ZtmDZtGvT09KSMSEREEnFyssbs2Y44fPge9uxxQ/v29aSORESVgFrNJBsbGxd5KbmUlBSYmBT/E3/B7HPnzp0Vxh0dHZGTk4NHjx6pNigREamty5ejIZcLCmNLl/bG1as+LMhEVGJqVZKtra0LrT1OS0tDfHx8obXKr2ratOkbj5udna2SfEREpL6ysvIwa9bP6NRpC/z9/1B4TldXC4aGb7/ePhFRAbUqyd27d0d4eDhSU1PFsdDQUGhoaKBr167F7le/fn3Y2toiPDxcYTw8PBx6enpvLdFERFS53boVBweHLViz5hIA4JNPTuHBg0SJUxFRZaZ0SX769CkWLFgAZ2dndOrUCZcvXwYAJCYmYunSpbh9+3apjzlmzBgYGhpi2rRpuHDhAg4ePIhVq1ZhzJgxCtdI9vb2Rt++fRX2nTlzJs6cOYOvvvoKFy9exIYNG7Bt2zaMHz+et80mIqqiBEGAn9/vaN9+E27ciAUA6OpqYuXKPrwxCBGViVIn7j148AAeHh6Qy+Vo1aoVHj9+jLy8PACAmZkZ/vzzT2RmZmLZsmWlOq6JiQl27NiBJUuWYNq0aTA0NMSIESMwc+ZMhe3kcjny8/MVxnr37o1vv/0WP/zwA/bu3YvatWvjww8/hI+PjzJfIhERqblnz9IxceIhnDjxQBxr0aI29uxxQ8uWxd98ioioJJQqyV9//TVq1KiB/fv3A3h5VYlX9ejRAydOnFAqkI2NDQICAt64za5du4ocHzBgAAYMGKDU6xIRUeVx9Og9TJx4CPHxmeLYjBkOWLGiD/T0qsSFm4hIYkr9TXL58mVMmzYNZmZmSEpKKvR8vXr1EBsbW+ZwREREr9u16y94eYWIj+vWNUJAgCucnXn+CRGpjlJrkgVBeON1hxMTE6Gjw7OIiYhI9Vxd7WFlZQoAGDLEDjduTGFBJiKVU6okN2vWDOfOnSvyuby8PBw7dgytW7cuUzAiIqKiGBvrIjDQDRs2DERIyGhYWBhKHYmIqiClSrKPjw/Onz+PL7/8Evfv3wcAJCQkIDw8HBMnTkRkZCRPmCMiojKLikqFm9uPePw4RWHc0bEhfH07QCaTSZSMiKo6pdYk9+jRA8uXL8eyZcvEk/fmzJkDQRBgZGSElStXomPHjioNSkRE1ctPP92Gj88RJCVl4fnzTPz6qzc0NdXq8v5EVIUpfQrw0KFD0a9fP4SHh+Pff/+FXC5Ho0aN0K1bNxgZGakyIxERVSNpadmYMSMU27dfF8cePkzG48cpsLLitY+JqGIoVZIFQYBMJoOBgQH69Omj6kxERFRNXboUBU/PIERE/HflpJEjm2HjxkGoWVNfwmREVN0o9Xurd999F0uXLsWff/6p6jxERFQN5eXJsXjxOXTrtk0syEZGOggIcMWPP45gQSaiCqfUTHKnTp1w8OBBBAYGok6dOnBxcYGLiwtatWql6nzSu3sACF8AZMRInYSIqEp6+DAJnp7BCA9/Io517twAu3cPg42NmYTJiKg6U6okf/vtt8jKysKvv/6KEydOYO/evQgICED9+vUxYMAAuLi44J133lF1VmmELwAS7/z3WKeGdFmIiKqgW7fixYKsoSHDggXdMX9+d2hp8SQ9IpKOTBAEoawHyczMxJkzZ3D8+HFcuHABubm5aNy4MUJDQ1WRURJOTk4AgNOj7gLp0YBMA6hpC3RdAtiOkDgdEVHVMm3aMZw48QC7d7uhS5eGUschokpC7GunT6v82CopyQUyMjIQHByMNWvWIDMzE//884+qDl3hCpVko/qAb5TEqYiIKr+//45D8+YWCtc4zszMRV6eHMbGuhImI6LKpjxLstKXgCvw4sULnDlzBidOnMD58+eRk5ODRo0awcXFRRX5iIioisjNzcfChWexfPkFbN48GJMmtROfMzDQljAZEVFhSpXk7OxsnD17FsePH0dYWBhevHiB+vXrY9y4cRgwYACaNWum6pxERFSJ3buXAA+PIFy58hQAMH16KJycrNGkiam0wYiIiqFUSe7cuTOysrJQu3ZtjBo1CgMGDEDr1q1VnY2IiCo5QRCwdes1zJgRiszMXACAlpYGvviiOxo2NJY4HRFR8ZQqyW5ubnBxcUGHDh1UnYeIiKqIhIRMTJ58BMHB/10hyNbWHHv2uKF9+3oSJiMiejulSvIXX3yh6hxERFSFnDoVAW/vEMTEpItjvr7tsXp1Pxga6kiYjIioZEpUki9fvgwA6Nixo8LjtynYnoiIqo/du29g3Lhg8XGtWgbYsmUwXF3tJUxFRFQ6JSrJ48aNg0wmw19//QUdHR3xcXEEQYBMJqvUl4AjIiLlDBjwP9SvXwPR0WlwdrbB9u2usLTkjZiIqHIpUUneuXMnAEBHR0fhMRER0evMzPSxc+cw3LwZiw8/dICGRvGTKkRE6qpEJblTp05vfExERNXTs2fp+Pjjk/jmm74Ks8W9e1uhd28rCZMREZWNhjI7eXl54bfffiv2+UuXLsHLy0vpUEREpP6OHr2HVq3WY8+em5gw4RDkcpXdwJWISHJKleQ//vgDz58/L/b5xMTEEp/cR0RElUtmZi6mTj2GwYP3Ij4+EwBw/fozPH6cInEyIiLVUfq21G86ce/Ro0cwNDRU9tBERKSmrl2Lgbt7EO7c+W+iZMgQO2zZMhgWFvx7n4iqjhKX5ODgYAQH/3dJn/Xr12P//v2FtktLS8Pdu3fRvXt31SQkIiLJyeUCVq8Ox/z5Z5CbKwcA6OtrYc0aZ/j4tH/jxAkRUWVU4pL84sULJCUliY8zMjKgoVF4tYaBgQHGjBmDadOmqSYhERFJKioqFd7eIThz5qE41q6dJQID3WBvX0vCZERE5afEJdnd3R3u7u4AgN69e2P+/PlwcnIqt2BERKQefvvtiViQZTLgk0+6YvHiXtDR0ZQ4GRFR+VFqTfKZM2dUnYOIiNTUyJHN4e19H6dPP8TOnUPRqxcv7UZEVV+JSvLTp08BAPXq1VN4/DYF2xMRUeXx4EEimjY1Uxjz83NBbq4cZmb6EqUiIqpYJSrJvXv3VrgtdcHjt+FtqYmIKo+8PDmWLTuPxYvPYdeuYRg7tqX4XI0auhImIyKqeCUqycuWLYNMJoO2trbCYyIiqhoePkyCp2cwwsOfAADef/8YunVrhIYNTSRORkQkjRKVZDc3tzc+JiKiykkQBAQG3sTUqceQlpYDANDQkOGjjzor3GaaiKi6UfpmIkXJyclBXl4eDAwMVHlYIiIqB8nJWXj//WPYt+9vcczKyhS7d7uhS5eGEiYjIpKeUrelPnbsGJYtW6Yw5u/vj3bt2qFjx46YNm0aMjIyVBKQiIhULyzsEVq33qBQkL28WuP69SksyEREULIkb9u2DS9evBAfX716Ff7+/ujWrRu8vb1x/vx5bNiwQWUhiYhIdQIDb6BnzwA8fpwCADA11cO+fcOxY8dQGBvzBD0iIkDJ5RZPnjzBsGHDxMdHjx5FrVq14O/vDy0tLQiCgJMnT+Ljjz9WWVAiIlKNPn2sUauWAeLjM9GjR2Ps3DkMjRrxBD0iolcpVZJzcnKgq/vfbMPFixfRvXt3aGm9PJyNjQ327NmjmoRERKRSdeoYYft2V9y8GYc5c7pAU1OpXyoSEVVpSv3N2KBBA4SHhwMAbt68iUePHuHdd98Vn09ISODJe0REauD580xMnnwYz59nKowPHGiLuXO7sSATERVDqZnk0aNH46uvvsKDBw8QGxuLunXrolevXuLzV69eRdOmTVUWkoiISu/kyQiMHx+CmJh0PH/+AkFBo3iNeyKiElKqJI8bNw66uro4d+4cWrRogffeew96enoAgOTkZMTHx2Ps2LEqDUpERCWTlZWHzz47jTVrLoljFy48xuPHKWjc2FS6YERElYjS10keNWoURo0aVWjc1NQUQUFBZQpFRETKuXUrDu7uQbhxI1Yc69fPBgEBrrw5CBFRKZT5ZiIPHjxAdHQ0AKB+/fpcZkFEJAFBEODv/wfmzDmF7Ox8AICuriZWruyDDz90gIYGl1kQEZWG0iX5l19+wYoVK8SCXKBBgwaYO3cunJycyhyOiIje7tmzdEyYcAihoQ/EsRYtamPPHje0bFlHwmRERJWXUiX53LlzmD59OurVq4eZM2fCxsYGABAREYH9+/fjww8/xIYNG9C9e3eVhiUiosJOnYpQKMgzZjhgxYo+0NMr8y8LiYiqLaX+Bv3hhx9gZ2eHwMBAhUu9OTk5wdPTE+7u7li3bh1LMhFRBfD0bIUjR+7h/PnHCAhwhbMzl70REZWVUhfIvHv3LoYOHVrktZANDAwwbNgw3L17t8zhiIiosKioVIXHMpkMGzcOwo0bU1iQiYhURKmSrKuri5SUlGKfT0lJUbgjX6WVmwGkR799OyKiCiCXC/j664uwtv4ehw7dUXiuZk19WFgYSpSMiKjqUaokOzg4YOfOnbh27Vqh5/766y/s2rULjo6OZQ4nuayk//5bh5dOIiLpREWlom/fXfjkk1+QmyvHpEmHEROTJnUsIqIqS6k1yXPmzMGYMWPg7u6OVq1awcrKCgDw8OFD3LhxA+bm5pg9e7ZKg0pD/t9/dl0iXQwiqtZ++uk2fHyOICkpCwAgkwGTJrWFuXnhJW9ERKQaSpXkhg0b4vDhw9i4cSPCwsJw/PhxAEC9evXg5eUFHx8fmJubqzSopIzqA7YjpE5BRNVMWlo2ZswIxfbt18WxBg2MsXPnUPTqZSVdMCKiaqDUJTk/Px+JiYkwNjbGZ599hs8++6w8chERVWuXLkXB0zMIERH/LfsaObIZNm4chJo19SVMRkRUPZS4JAuCgDVr1mD37t148eIFNDU10aNHD3z11VcwNTUtx4hERNXLnj034eUVjPx8AQBgZKQDf38XeHm1hkzGO+cREVWEEpfkoKAgbNq0CXXr1sW7776LJ0+e4PTp05DL5Vi/fn15ZiQiqla6d28MY2NdJCVloXPnBti9exhsbMykjkVEVK2UuCTv3bsXzZo1w549e6CnpwcAWLp0Kfbs2YPExESYmfEvcCIiVWjQwBibNg3GrVtxmD+/O7S0lLoQERERlUGJ/+Z98uQJXF1dxYIMAO7u7pDL5Xj06FG5hCMiquqSk7MwY8YJpKRkKYyPGNEMX37ZkwWZiEgiJZ5JTklJKTRbXLNmTQBAdna2alMREVUDYWGPMG5cMB4/TkFCwgvs3u0mdSQiIvp/pZqi4AkjRERll5ubj/nzT6NnzwA8fvzy7qVHj97DkyfF38mUiIgqVqkuAbd69Wps3LhRfCyXv7zZxueffw59fcVLEslkMhw+fFgFEYmIqo579xLg4RGEK1eeimM9ejTGzp3D0LChiYTJiIjoVSUuyR07dixynCfsERG9nSAI2Lr1GmbMCEVmZi4AQEtLA0uW9MKcOV2gqcm1x0RE6qTEJXnXrl3lmYOIqMpKSMjE5MlHEBx8RxyztTXHnj1uaN++noTJiIioOErdlpqIiEouKOgfhYLs49MO337rDENDHQlTERHRm/D3e0RE5ey999phwID/wdxcHyEho7Fx42AWZCIiNceZZCIiFYuLy0Dt2obiY5lMhu3bXZGXJ0e9ejUkTEZERCXFmWQiIhURBAF+fr+jSZPvcPJkhMJztWsbsiATEVUiLMlERCrw7Fk6BgzYg+nTQ/HiRR7Gjw/B8+eZUsciIiIlcbkFEVEZHT16DxMnHkJ8/H+leNSo5jAy4rpjIqLKqkwlOTY2FpcvX0ZCQgKcnZ1Rt25d5OfnIy0tDTVq1ICmpqaqchIRqZ3MzFzMnn0S69dfEcfq1jVCQIArnJ2bSpiMiIjKSqmSLAgCVqxYgcDAQOTl5UEmk8HW1hZ169ZFZmYmevfujenTp2P8+PEqjktEpB6uXo2Bh0cQ7tx5Lo4NGWKHLVsGw8LC8A17EhFRZaDUmuQtW7Zg586dmDhxIrZv3w5BEMTnatSogX79+uHkyZMqC0lEpE727r2Jzp23iAVZX18LGzYMREjIaBZkIqIqQqmZ5AMHDmDo0KGYNWsWkpKSCj1vZ2eHsLCwMocjIlJHnTrVh66uFnJzc9CunSUCA91gb19L6lhERKRCSpXkmJgYtG3bttjn9fX1kZ6ernQoIiJ1ZmNjhnXrBuD27XgsXtwLOjo8/4KIqKpRqiSbm5sjJiam2Odv3boFS0tLpUMREamLtLRsLFt2HvPnd1e4WoWXV2sJUxERUXlTak1y3759sW/fPjx58kQck8lkAIALFy4gODgY/fv3V01CIiKJXLoUhbZtN2LFiouYNetnqeMQEVEFUqokT58+HRYWFnB1dcWnn34KmUyGzZs3Y+zYsZg8eTJsbW0xZcoUVWclIqoQeXlyLF58Dt26bUNExMvzLvbt+xvR0akSJyMiooqiVEmuUaMG9u/fj/feew+xsbHQ1dXF5cuXkZaWhmnTpmHPnj3Q19dXdVYionL38GESevQIwJdfnkV+/ssr93Tu3ADXrvmifn1jidMREVFFUfpmInp6epg6dSqmTp2qyjyIiIjA0qVLce3aNRgaGsLV1RUfffQRdHRKfueqgIAALF++HD179sTGjRtVmo+IqiZBEBAYeBNTpx5DWloOAEBDQ4YFC7pj/vzu0NJSak6BiIgqKbW6LXVKSgq8vb3RpEkT+Pn5ITY2FitWrEBWVhYWLFhQomPEx8dj3bp1MDc3L+e0RFRVJCdn4f33j2Hfvr/FMSsrUwQGusHRsaGEyYiISCpKleR58+a9dRuZTIZly5aV6rj79u1DRkYG/P39YWpqCgDIz8/HokWL4Ovrizp16rz1GF9//TV69+6Np0+fluq1iaj6Cgy8oVCQvb1bY+1aFxgb60qYioiIpKRUSf79998LjcnlcsTHxyM/Px9mZmZKrUkOCwuDo6OjWJABwMXFBV9++SUuXrwINze3N+5/5coV/PLLLwgNDcXHH39c6tcnourp/fc74uDBf3Dt2jNs2DAQo0e3kDoSERFJTKmSfObMmSLHc3Nz8eOPP2LHjh3Ytm1bqY8bGRmJ4cOHK4wZGxvDwsICkZGRb9w3Pz8fS5YswZQpU1C7du1SvzYRVR/JyVkwNdUTH2toyLBr1zDk5wto1MhEwmRERKQuVHomira2Njw9PdG1a1csWbKk1PunpqbC2Ljw2eMmJiZISUl547579uzBixcvMH78+FK/LhFVD4IgYMuWq2jUaA3On3+k8Fz9+sYsyEREJCqX07Xt7e1x+fLl8jh0kRISErB27VrMnTu3VFfBIKLqIyEhE8OH78fkyUeQlpYDT89gJCdnSR2LiIjUVLlc3SI8PFypNcnGxsZIS0srNJ6SkgITk+JneL7//nvY2dmhQ4cOSE19ebH/vLw85OXlITU1FQYGBtDSUqsLeRBRBTp1KgLe3iGIiUkXx/r3t4G2Ni/rRkRERVOqOfr7+xc5npaWhsuXL+P27dvw8fEp9XGtra0LrT1OS0tDfHw8rK2ti93v4cOHuHz5Mjp27FjouY4dO2Lz5s3o3r17qfMQUeWWlZWHzz47jTVrLolj5ub62Lp1CFxd7SVMRkRE6k6lJdnExAQNGzbEokWLMGrUqFIft3v37tiwYYPC2uTQ0FBoaGiga9euxe732WefiTPIBZYtWwY9PT3MmjULdnZ2pc4CAJDnKbcfEUnu1q04uLsH4caNWHGsXz8bBAS4wtKyhoTJiIioMlCqJN+5c0fVOQAAY8aMwa5duzBt2jT4+voiNjYWq1atwpgxYxSukezt7Y2nT5/i1KlTAIB33nmn0LGMjY1hYGAABweHsgfT4T+oRJXJjz/+DW/vEGRn5wMAdHQ0sXJlH0yf7gANDZnE6YiIqDIo9YK8rKwsLF++vNjLwJWFiYkJduzYAU1NTUybNg2rV6/GiBEjMHfuXIXt5HI58vPzVf76xepa+it1EJF0WrWqA5nsZRlu0aI2Ll+ejI8+6syCTEREJVbqmWQ9PT38+OOPaNq0aXnkgY2NDQICAt64za5du956nJJsUyJG9QHbEao5FhFViHfescC33/bD3bsJWLGiD/T0eOIuERGVjlL/cjRv3hz37t1TdRYiolLLzMzFt9/+htmzuyiU4fffL3wiLxERUUkpdf2jzz77DMePH8eBAweQl8eT24hIGteuxaB9+0344otf8dlnp6WOQ0REVYhMEAShJBtevnwZNjY2MDMzw+DBg5GUlISEhATo6OigTp060NXVVTywTIbDhw+XS+iK4OTkBKRE4vTkXMA3Suo4RPQKuVzA6tXhmD//DHJz5QAAAwNtPHjwIa9cQURUjTg5OQEATp9W/URJiZdbeHl54euvv8agQYNgamoKU1NTWFlZqTwQEdGbREWlwts7BGfOPBTH2rWzRGCgGwsyERGpTIlLsiAIKJh0VtlJcUREpfDTT7fh43MESUkvbyctkwGffNIVixf3go6OpsTpiIioKuEp30Sk9tLSsjFjRii2b78ujjVoYIydO4eiVy/+RouIiFSvVCW54LqjREQVacuWqwoFeeTIZti4cRBq1tSXLhQREVVppSrJc+bMwZw5c0q0rUwmw+3bt5UKRUT0qg8/dMD+/bfx999x8Pd3gZdXa/7QTkRE5apUJblLly5o0qRJOUUhInopIyMHhoY64mMtLQ3s2eMGuVyAjY2ZhMmIiKi6KFVJHjp0KAYPHlxeWYiomhMEAYGBN/HRR6EIDfVEhw71xOesrGpKmIyIiKobpW4mQkSkasnJWXB3D8K4ccFISHgBD48gZGTkSB2LiIiqKV7dgogkFxb2COPGBePx4xRxrHPnBpDLS3SvIyIiIpVjSSYiyeTm5mPhwrNYvvwCCu79aWqqhw0bBmL06BbShiMiomqtxCX5zp075ZmDiKqZe/cS4OERhCtXnopjPXo0xs6dw9CokYmEyYiIiDiTTEQS+Omn2/D2DkFmZi6Al1evWLKkF+bM6QJNTZ4qQURE0mNJJqIKZ2NTE7m5+QAAW1tz7Nnjhvbt671lLyIioorDkkxEFa5tW0ssW+aE+/cT8O23zgrXRCYiIlIHLMlEVK6ysvLwww+X8eGHnaCtrSmOf/yxI++aR0REaoslmYjKzd9/x8Hd/SBu3oxDYuILLF3aW3yOBZmIiNQZz5AhIpUTBAF+fr+jQ4dNuHkzDgDw7be/ITY2XeJkREREJcOZZCJSqWfP0jFhwiGEhj4Qx1q0qI09e9xQp46RhMmIiIhKjiWZiFTm6NF7mDjxEOLjM8WxGTMcsGJFH+jp8a8bIiKqPPivFhGVWWZmLmbPPon166+IY3XrGiEgwBXOzk0lTEZERKQclmQiKjN//z8UCvKQIXbYsmUwLCwMJUxFRESkPJ64R0RlNmOGA9q0qQt9fS1s2DAQISGjWZCJiKhS40wyEZVadnYedHX/++tDV1cL+/YNhyAA9va1JExGRESkGpxJJqJS+emn27Cy+h43b8YqjNvZ1WJBJiKiKoMlmYhKJC0tGxMnHsLIkQcQE5MOd/cgZGXlSR2LiIioXHC5BRG91aVLUfD0DEJERJI41qyZBXJy8nlpNyIiqpL4rxsRFSsvT45ly85j8eJzyM8XAABGRjpYt24Axo1rxVtLExFRlcWSTERFevgwCZ6ewQgPfyKOOTo2wO7dbrC2rilhMiIiovLHkkxEhYSE3IGXVzDS0nIAABoaMixY0B3z53eHlhZPZSAioqqPJZmICqlTxxCZmbkAACsrU+ze7YYuXRpKnIqIiKjisCQTUSGOjg2xYEEPREQkwc/PBcbGulJHIiIiqlAsyUTVXG5uPrZtu4b33msHTc3/llJ88UV3nphHRETVFksyUTV2714CPDyCcOXKUyQmvsC8ee+Kz7EgExFRdcYzcIiqIUEQsGXLVbRtuxFXrjwFACxZEob4+AyJkxEREakHziQTVTMJCZmYPPkIgoPviGO2tuYIDHSDhYWhhMmIiIjUB0syUTVy6lQEvL1DEBOTLo75+LTDt986w9BQR8JkRERE6oUlmagayM7Ow2efnca3314Sx8zN9bF16xC4utpLmIyIiEg9sSQTVQOrV/+mUJD79bNBQIArLC1rSJiKiIhIffHEPaJq4KOPOuOdd2pBV1cT333njBMnPFiQiYiI3oAzyURVUF6eXOH20QYG2vjxxxEAgJYt60gVi4iIqNLgTDJRFXP06D38739+uHcvQWG8Zcs6LMhEREQlxJJMVEVkZuZi6tRjGDx4L/79NxkeHkHIzc2XOhYREVGlxOUWRFXAtWsxcHcPwp07z8WxevVqIDMzFyYmmhImIyIiqpxYkokqMblcwOrV4Zg//wxyc+UAAH19LaxZ4wwfn/a8tTQREZGSWJKJKqmoqFR4e4fgzJmH4li7dpYIDHSDvX0tCZMRERFVfizJRJXQ0aP34OUVjKSkLACATAZ88klXLF7cCzo6XF5BRERUVizJRJWQoaE2kpNfFuQGDYyxc+dQ9OplJXEqIiKiqoMlmagS6tXLCnPmdMHDh8nYuHEQatbUlzoSERFRlcKSTKTm8vLk2Lv3Jjw8WkFD478T8ZYtc4KGhown5xEREZUDlmQiNfbwYRI8PYMRHv4EiYkvMGNGZ/E5TU1e5pyIiKi88F9ZIjUkCAJ27foLrVtvQHj4EwDAZ5+dwfPnmRInIyIiqh44k0ykZpKTs/D++8ewb9/f4piVlSl273ZDrVoGEiYjIiKqPliSidRIWNgjjBsXjMePU8QxL6/W8PNzgbGxroTJiIiIqheWZCI1kJOTj4ULz2LFigsQhJdjJia62LhxEEaPbiFtOCIiomqIJZlIDaxYcQHLl18QH/fo0Rg7dw5Do0YmEqYiIiKqvnjiHpEa+OijzrC2rgktLQ0sX+6E06e9WJCJiIgkxJlkIgnI5YLCNY+NjXWxb99waGjI0L59PQmTEREREcCZZKIKd+pUBJo3/wH//pusMN6xY30WZCIiIjXBkkxUQbKy8jBr1s/o12837tx5Di+vYOTny6WORUREREXgcguiCnDrVhzc3YNw40asOGZgoI20tByYmupJmIyIiIiKwplkonIkCAL8/H5H+/abxIKsq6uJ77/vj+PHPViQiYiI1BRnkonKybNn6Zgw4RBCQx+IYy1a1MaePW5o2bKOhMmIiIjobViSicpBaOgDeHkFIz4+UxybMcMBK1b0gZ4ev+2IiIjUHf+1JioHcrkgFuS6dY0QEOAKZ+emEqciIiKikmJJJioHAwb8D9OmdcSTJ6nYsmUwLCwMpY5EREREpcCSTFRGcrmAkJA7GDbMHjLZfzcIWbPGGVpaGgpjREREVDnw6hZEZRAVlYq+fXdh+PD92Lr1msJz2tqaLMhERESVFEsykZIOHLiFVq3W48yZhwCAmTN/RmLiC4lTERERkSpwuQVRKaWlZWP69FAEBFwXxxo0MMbOnUNhZqYvXTAiIiJSGZZkolK4dCkKHh5BiIxMEsdGjmyGjRsHoWZNFmQiIqKqgiWZqATy8uRYtuw8Fi8+h/x8AQBgZKQDf38XeHm15tpjIiKiKoYlmagEliw5h8WLw8THnTs3wO7dw2BjYyZhKiIiIiovanfiXkREBCZMmIA2bdqga9euWLVqFXJyct64T1xcHFatWgVXV1e0bdsW3bt3x8cff4zo6OgKSk1V3YwZndGggTE0NGRYuLAHzp+fwIJMRERUhanVTHJKSgq8vb3RpEkT+Pn5ITY2FitWrEBWVhYWLFhQ7H63bt3CqVOnMHz4cLRu3RpJSUlYv349Ro4ciaNHj8LMjGWGSkcQBIUlFGZm+ti7dzg0NGTo0qWhhMmIiIioIqhVSd63bx8yMjLg7+8PU1NTAEB+fj4WLVoEX19f1KlTp8j92rdvjxMnTkBL678vp127dujZsydCQkIwceLEiohPVURY2CPMnPkzjh4dC0vLGuJ4t26NJExFREREFUmtlluEhYXB0dFRLMgA4OLiArlcjosXLxa7n7GxsUJBBoC6devCzMwMcXFx5RWXqpjc3HzMn38aPXsG4OrVGIwffwhyuSB1LCIiIpKAWpXkyMhIWFtbK4wZGxvDwsICkZGRpTrWw4cPkZCQABsbG1VGpCrq3r0EdOmyDcuWXYDw/704OzsPqanZ0gYjIiIiSahVSU5NTYWxsXGhcRMTE6SkpJT4OIIgYOnSpahduzYGDhyoyohUxQiCgC1brqJt2424cuUpAEBLSwPLlzvh9GkvmJrqSZyQiIiIpKBWa5JVxc/PD5cuXcKWLVtgYGAgdRxSUwkJmZg8+QiCg++IY7a25tizxw3t29eTMBkRERFJTa1KsrGxMdLS0gqNp6SkwMTEpETH2L9/P9atW4evvvoKjo6Oqo5IVcTp05EYNy4YMTHp4pivb3usXt0PhoY6EiYjIiIidaBWJdna2rrQ2uO0tDTEx8cXWqtclFOnTmHhwoWYPn06RowYUV4xqQpITHwhFmRzc31s3ToErq72EqciIiIidaFWJbl79+7YsGGDwtrk0NBQaGhooGvXrm/c9/fff8esWbMwcuRITJs2rSLiUiU2cmRzjB//AE+fpiEgwFXhUm9EREREalWSx4wZg127dmHatGnw9fVFbGwsVq1ahTFjxihcI9nb2xtPnz7FqVOnALy8S9+0adPQpEkTuLq64vr16+K2ZmZmaNSI17etzgRBwM8/R6B//6YK4+vXD4SOjiY0NGTF7ElERETVlVqVZBMTE+zYsQNLlizBtGnTYGhoiBEjRmDmzJkK28nlcuTn54uP//rrL6SlpSEtLQ1jx45V2HbYsGFYsWJFheQn9fPsWTomTjyEEyceIDDQDe7uLcXn9PTU6o8/ERERqRGZIAi8W0IRnJycgJRInJ6cC/hGSR2HlHD06D1MnHgI8fGZAABTUz08fDiDl3UjIiKqIpycnAAAp0+fVvmxOZVGVU5mZi5mzz6J9euviGN16xohIMCVBZmIiIhKhCWZqpRr12Lg7h6EO3eei2NDhthhy5bBsLAwlDAZERERVSYsyVQlyOUCVq8Ox/z5Z5CbKwcA6OtrYc0aZ/j4tIdMxpPziIiIqORYkqlKWLjwLJYsCRMft2tnicBAN9jb15IwFREREVVWGlIHIFKFDz7ohNq1DSGTAZ9+2hW//TaJBZmIiIiUxplkqhJq1zbE7t3DoKWlgV69rKSOQ0RERJUcZ5Kp0rl0KQrvvrsdz59nKoz37WvDgkxEREQqwZJMlUZenhyLF59Dt27bcOHCY7z33mHwMt9ERERUHrjcgiqFhw+T4OkZjPDwJ+JYbGwGUlOzYWLCax8TERGRanEmmdSaIAjYvfsGWrfeIBZkDQ0ZFi7sgfPnJ7AgExERUbngTDKpreTkLLz//jHs2/e3OGZlZYrdu93QpUtDCZMRERFRVceSTGrp/PlH8PQMxuPHKeKYl1dr+Pm5wNhYV8JkREREVB2wJJNaioxMEguyqakeNmwYiNGjW0icioiIiKoLlmRSS15erXH8+APExqZj165haNjQROpIREREVI2wJJPkBEHAhQuP8e67jcUxmUyGbduGQE9PC5qaPL+UiIiIKhbbB0kqISETw4fvR/fuAQgJuaPwnKGhDgsyERERSYINhCRz6lQEWrZcj+Dgl+X4vfcOIzU1W+JURERERCzJJIGsrDzMmvUz+vXbjZiYdACAubk+tm4dwitXEBERkVrgmmSqULduxcHdPQg3bsSKY/362SAgwBWWljUkTEZERET0H5ZkqhCCIMDf/w/MmXMK2dn5AABdXU2sXNkHH37oAA0NmcQJiYiIiP7DkkwV4ssvz2LJkjDxcYsWtbFnjxtatqwjYSoiIiKionFNMlUIH5/2qFlTDwAwY4YDLl+ezIJMREREaoszyVQhGjQwxo4dQ6GtrYn+/ZtKHYeIiIjojTiTTCp37VoM+vffjeTkLIXxwYPtWJCJiIioUmBJJpWRywV8/fVFODhswc8/R2DatONSRyIiIiJSCpdbkEpERaXC2zsEZ848FMf++SceqanZvPYxERERVTqcSaYy++mn22jVar1YkGUy4JNPuuDSpfdYkImIiKhS4kwyKS0tLRszZoRi+/br4liDBsbYuXMoevWyki4YERERURmxJJNSfv89Ch4eQYiISBLHRo5sho0bB6FmTX0JkxERERGVHUsyKeXq1RixIBsZ6cDf3wVeXq0hk/HOeURERFT5sSSTUqZM6YDjxx/g+fNM7N49DDY2ZlJHIiIiIlIZlmR6K0EQ8OefMejQoZ44JpPJsHv3MBga6kBLi+d/EhERUdXCdkNvlJycBXf3IHTsuBk///xA4TkTEz0WZCIiIqqS2HCoWGFhj9C69Qbs2/c3AGD8+ENIS8uWOBURERFR+WNJpkJyc/Mxf/5p9OwZgMePUwAApqZ6+O47Z9SoweseExERUdXHNcmk4N69BHh4BOHKlafiWI8ejbFz5zA0amQiYTIiIiKiisOSTABenpy3ZctVfPTRz8jMzAUAaGlpYMmSXpgzpws0NflLByIiIqo+WJIJALBw4VksXhwmPra1NceePW5o377eG/YiIiIiqpo4PUgAAG/vNqhRQwcA4OPTDlev+rAgExERUbXFmWQCAFhb18TmzYOhq6uFoUPtpY5DREREJCnOJFdDt27FYfjw/UhPz1EYHz26BQsyERERETiTXK0IggB//z8wZ84pZGfnw8xMD5s3D5E6FhEREZHaYUmuJp49S8eECYcQGvrfXfMuXYpGWlo2r31MRERE9Bout6gGjhy5i5Yt1ysU5BkzHHD58mQWZCIiIqIicCa5CsvMzMXs2Sexfv0VcaxuXSMEBLjC2bmphMmIiIiI1BtLchV19WoMPDyCcOfOc3FsyBA7bNkyGBYWhhImIyKqWPn5+cjNzZU6BhGVkra2NjQ1NSV7fZbkKurMmYdiQdbX18KaNc7w8WkPmUwmcTIiooohCAKePXuG5ORkqaMQkZJMTU1Rt25dSfoLS3IVNWuWI06ceIDk5CwEBrrB3r6W1JGIiCpUQUGuXbs2DAwMOElAVIkIgoDMzEzExcUBACwtLSs8A0tyFXH7djyaNbMQH2toyLB//wjUqKELHR3pflVBRCSF/Px8sSCbm5tLHYeIlKCvrw8AiIuLQ+3atSt86QWvblHJpafnYNKkQ2jZcj3Cwh4pPGdubsCCTETVUsEaZAMDA4mTEFFZFHwPS3FeAUtyJfb771Fo02YDtm27DrlcwLhxwYXuokdEVJ1xiQVR5Sbl9zBLciWUlyfHkiXn0LXrNkREJAEAjIx0sHhxTxgaakucjoiIiKjy45rkSubhwySMGxeMixefiGOdOzfA7t3DYGNjJmEyIiIioqqDM8mVhCAI2L37Blq33iAWZA0NGb78sgfOn5/AgkxEVEX5+fnBzs5O/J+DgwPGjh2Lc+fOFbl9SkoKVq5ciT59+qBFixbo0qULZs2ahYiIiCK3z8jIgL+/PwYNGoTWrVujTZs2GDFiBLZv347s7Ozy/NLUwogRIxAYGCh1jAp34MABODs7o2XLlhgyZAh+/fXXEu137do1uLu7o1WrVujSpQuWLFmCFy9eKGxz8+ZNzJs3Dy4uLrC3t4evr2+h40RFRaFNmzaIiopSyddTHjiTXEksWnQOixb99xeilZUpdu92Q5cuDSVMRUREFUFPTw87duwA8PJM/w0bNmDKlCkIDAxEu3btxO3i4+Ph6emJlJQUTJkyBc2aNcOzZ8+wbds2jBgxAps2bULHjh3F7RMTE+Ht7Y2YmBh4e3ujffv2AF4WoU2bNkFDQwPe3t4V+8VWoFOnTiE6OhrDhw+XOkqFOnbsGL744gtMmTIFnTt3xvHjx/HBBx8gMDAQbdq0KXa/6OhojB8/Hh06dICfnx/i4uLwzTffID4+HmvXrhW3u3r1Kq5cuYJWrVoV+4NWgwYN4OzsDD8/P6xcuVLVX6JKsCRXEqNHN8eqVRfx4kUevLxaw8/PBcbGulLHIiKiCqChoaFQXlq3bo0ePXogJCREoSQvWrQIT58+RUhICGxsbMTxPn36YMSIEfj4449x6tQp6Orqits/efIE+/fvh62trbh9ly5d4OHhgcjIyPL/4oqQlZUFPT29cn+dHTt2YODAgSp5rYrKrApr167FwIED8dFHHwEAOnfujHv37mHdunXYvHlzsftt3LgRxsbGWL9+PXR0dAAAxsbGmD59Om7fvo1mzZoBAMaNGyf+cDVu3LhijzdixAhMmDABn376KczM1O834lxuUUm8844FfvhhIPbtG44dO4ayIBMRVWN16tSBmZkZnj59Ko5FR0fjl19+wdChQxUKMvDyMlpTpkxBbGwsTpw4IW7/888/Y8yYMQoFuYCpqalCAS9KREQEPvjgA3Tq1AmtW7fGkCFDcPToUQAvf51uZ2eH0NBQhX2++uor9O7dW3wcFBQEOzs7XLt2DRMmTECbNm2watUqjBs3rshf0+/evRutWrVCWloagJfLEbdu3QpnZ2e0aNECTk5OCAgIeGNuAHjy5AmuXLmC/v37K4yfPXsWEyZMgKOjI9q1a4eRI0ciLCxMYZviMgMvb2Ize/ZsODg4oFWrVvDw8MDff/+tsH9ISAjGjh2LTp06oWPHjhg3bhxu3Ljx1syq8OTJE/z7779wcXFRGB8wYAB+++035OQUf5Wsf/75Bx07dhQLMgB069YNAHDmzBlxTEOjZPWyffv2MDU1xZEjR0rzJVQYziSroXv3ErB0aRg2bRoMPb3/PqLx49tIF4qIiNRGRkYGUlJS0KBBA3Hs8uXLEAQBvXr1KnKfgmJ65coVDB06FFeuXIEgCHj33XeVyvDvv/9i9OjRsLS0xPz582FhYYF79+4pFPfS+PjjjzF69Gj4+vpCX18f//zzD5YuXYrk5GSYmpqK2x09ehQ9evRAjRo1ALws3QcOHMCUKVPQunVrXL16Fd988w10dXUxduzYYl/v0qVL0NLSQqtWrRTGo6Ki0KtXL0ycOBEaGhoICwuDj48PduzYAQcHhzdmTklJgbu7OwwMDPDFF1+gRo0a2LVrF7y9vXHy5EnxxjZRUVEYOnQoGjVqhJycHBw7dgweHh44fPgwrKysis0sCALy8/Pf+l5qaRVf7wp+O/D669jY2CA3NxdPnjwp9ENWgezsbIWCDADa2tqQyWRK/dZBQ0MDrVu3Rnh4uFou62FJViMvfxq+hhkzQpGZmQszM318913/t+9IREQlc/cAEL4AyEmT5vV1agBdlwC2I0q9a15eHoCXa5K//vprGBoawsvLS3y+4Pa99erVK3J/IyMjGBsb49mzZwCA2NhYAMrf7tfPzw/a2trYu3cvjIyMALxcpqGsMWPGwMfHR3zcqFEjLF26FCdPnsSoUaMAvJz9vn79Or777jsAwOPHj7F7924sWrQIo0ePFjNkZWVh3bp1GD16dLGzmjdv3kSTJk0KlT5PT0/xv+VyORwcHPDgwQPs37+/UEl+PfPatWuRmpqKAwcOiIXY0dERzs7O2Lp1Kz755BMAwAcffKDwGl27dsWNGzcQHByMWbNmFfseBQcHY968ecU+X+D06dMKP0C9KiUlBcDLZRKvKnhc8HxRmjRpgps3b0IQBPH6xTdu3IAgCG/c703s7e3V9sRJlmQ1kZCQicmTjyA4+I44Fhr6ABkZOTA01HnDnkREVGJXvgYS77x9u/J0+etSl+TMzEw0b95cfKypqYkffvgB1tbWZY6j7M0aLl26BGdnZ7Egl1XPnj0VHtesWRNdunTBsWPHxJJ8/PhxGBgYiLPl4eHhAIB+/fqJP0QAL4vy5s2bERMTg/r16xf5enFxcahZs2ah8WfPnmHNmjUIDw9HfHw8BEEAAIX3v7jMFy9ehIODA0xMTMQ8Ghoa6NixI27evCluFxERgW+//RbXrl1DQkKCOP7vv/8WmbVAr1698NNPP71xGwCoXbv2W7dRxtixYzF+/HisXr0aEydORFxcHBYtWlSm20XXrFkTSUlJyM3Nhba2et3rgSVZDZw6FQFv7xDExKSLYz4+7fDtt84syEREqtTxE+DiF9LOJHecU+rd9PT0sHv3bgiCgH///RerV6/Gp59+iiNHjoiFqOD/nz59Cnt7+0LHSE9PR2pqKurWrQvg5bpmAIiJiXnjr/iLk5ycrNIyVqtWrUJjAwcOxNy5cxEfHw8LCwscO3YMffv2FU88TEpKgiAI6Ny5c5HHfFNJzsnJKTSLLJfL8f777yMtLQ3Tp09H48aNoa+vj7Vr1yImJuatmZOSknD9+vUiC3WjRo0AvPwcJk6cCDMzM8ydOxf16tWDrq4uPv/887decs/U1FRcZvImb1puYWJiAgBIS0uDhYWFOJ6amqrwfFEcHR0xe/Zs+Pv7Y/PmzdDQ0MCYMWOgra2t9J+Fgs8gOzubJZn+k5WVh88+O401ay6JY+bm+ti6dQhcXQv/BUdERGVkO0KppQ5S09DQQMuWLQEArVq1gpWVFUaNGoV169Zh0aJFAICOHTtCJpPh7NmzCifGFTh79iwAoEOHDgrbnz9/XqllEqampuISj6IUFNnc3FyF8YIyVhJOTk7Q0dHBiRMn0K1bN/zzzz8KyxFMTEwgk8mwZ8+eIgvWm8q/iYkJoqOjFcYePXqE27dvY926dejTp484npWVVaK8JiYmePfddzFjxoxCzxWUwevXr+PZs2fYuHGjwg8zaWlp4g8wxVHFcouC3z5ERkYq/CYiMjIS2traaNjwzZeWnTx5Mjw8PPDkyRNYWFjA2NgYnTt3Fmf7Sys1NRXa2toq+42EKrEkS+TWrTi4uwfhxo1YcaxfPxsEBLjC0vLtPyUSEVH11bJlSwwcOBBBQUH44IMPYGFhgfr166NPnz4ICQnBhAkTFAriixcvsGHDBtStW1e8qkG9evXg7OyMffv2Yfjw4WjatKnCa6SmpiIiIgJt27YtMoOjoyN+/vlnzJ49u8iCY25uDm1tbYWbmOTk5ODy5csl/jqNjIzQs2dPHDt2DCkpKTAzM1Mo9I6OjgBezmoX9YPBm1hZWeH3339XGCuYyX21cEdHR+PatWto0qTJW4/ZpUsXHD58GDY2NjAwMChym4LC/eprXL16FdHR0fjf//73xuOrYrlFw4YN0aRJE4SGhir8IHD8+HE4OjoWml0vioGBAezs7AAAP/30EwRBKHS1jJKKjo5W6jcZFYElWSJBQf+IBVlXVxMrV/bBhx86QENDubVhRERUvUydOhXHjx/Hjh07MHv2bADAl19+CU9PT3h4eMDX1xfNmjVDbGwstm3bhujoaGzatEmc4S3Y3svLC2PHjlW4mchff/2F3bt3Y/LkycWW5A8++ABnz56Fu7s73nvvPVhYWCAiIgIvXrzA5MmToaGhgb59+yIwMBCNGzdGzZo1xSUjpVkHPWjQIHzwwQeIjo5G//79FZYSWFlZwcPDA5988gkmTZqE1q1bIzc3F//++y9+//13/PDDD8Uet127dli3bh2ePXsmzuBaW1ujbt26WL16NeRyOTIzM7F27doSLyUYP348jhw5Ak9PT3h5eaFevXpITEzEX3/9hTp16mD8+PFo06YNDAwMsGjRIvj4+CA2NhZ+fn7i8pc3qVmzZpHrqEvrww8/xOzZs9GoUSM4ODjg+PHjuHHjBnbv3i1uEx0djb59+2Lq1KniiYZPnjxBSEiIeEWQS5cuYefOnVi2bJnCMo3ExET88ccf4n9nZGSIlwLs0aMH9PX1xW3//vtv8c+dumFJlsi8ee8iNDQCqanZ2LPHDS1bvv2bg4iIqIC1tTUGDBiAvXv3wtfXFzVq1ICFhQX279+PDRs2YOfOnYiNjUWNGjXQuXNnfP3114Uu7WVmZoZ9+/YhICAAJ06cEO+y17RpU7z33nsYM2ZMsa/fpEkT7Nu3D6tXr8aiRYuQn5+PJk2aKFzt4YsvvsAXX3yBpUuXwtDQEJMmTYKVlRVOnz5d4q+z4HJv8fHxGDhwYKHnP//8c1hZWeHHH3/EunXrYGhoCCsrq0LXP35dp06dYGpqirCwMHGpgI6ODvz8/LB48WLMmDEDlpaWeP/993Hp0qVC1zouSs2aNfHjjz/iu+++wzfffIPk5GSYm5ujdevW6Nu3L4CX65i///57rFq1ClOnTkWTJk2waNEibNmypcTvSVkNGjQIL168wObNm7Fp0yZYWVnB399f4QeigsvNFZy4CLyc/f7jjz+wY8cO5Obmwt7eHv7+/oUuO3j//v1CS04KHr+6FCQhIQG3bt164xU9pCQTXv3qSeTk5ASkROL05FzAt+z3FY+MTIK1teJPf8+epcPUVE/hWshERFR2WVlZePjwIaysrCrNXdCo4q1YsQK3b9/Gzp07pY5SLQUGBiIgIAAnT54s9rcLb/tednJyAoBS/eBVUrzjXjnLzMzF1KnHYGfnj8uXFU8QqFvXiAWZiIhIIhMnTsSNGzdw547ElwWshuRyOXbu3Ilp06YpfRnC8saSXI6uXYtB+/absH79FeTlyeHhEYTMzNy370hERETlrnbt2li+fDkSExOljlLtxMXFYdiwYRgyZIjUUYrFacxyIJcLWL06HPPnn0FurhwAoK+vhY8/doS+Pt9yIiIidaHsVRmobOrWrYspU6ZIHeON2NhULCoqFd7eIThz5qE41q6dJQID3WBvX/hC6URERESkfliSVeinn27Dx+cIkpJeXgNRJgM++aQrFi/uBR0d5W/ZSEREREQViyVZRRYvPocvvzwrPm7QwBi7dg1Dz55NJMtERFTd8QJORJWblN/DPHFPRYYMsYO29su3c+TIZrhxYwoLMhGRRAruZpaZmSlxEiIqi4Lv4aJuO17eOJOsIm3a1MWaNc4wMtKBl1drtb2cCRFRdaCpqQlTU1PExcUBeHkbXf69TFR5CIKAzMxMxMXFwdTUFJqaFb9slSVZCQ8fJmHVqotYu9YF2tr/fWjTpnWSMBUREb2q4FbDBUWZiCofU1NT8Xu5oqldSY6IiMDSpUtx7do1GBoawtXVFR999BF0dHTeuJ8gCNi8eTP27NmDxMREvPPOO5g3bx7atGmjsmyCICAw8CamTj2GtLQcmJnp46uvnFR2fCIiUh2ZTAZLS0vUrl0bubm8Rj1RZaOtrS3JDHIBtSrJKSkp8Pb2RpMmTeDn54fY2FisWLECWVlZWLBgwRv33bx5M9auXYvZs2fDzs4OgYGBmDhxIg4dOoSGDRuWOVtychbef/8Y9u37797t+/ffxvz53WFgUPHrZIiIqGQ0NTUl/YeWiContSrJ+/btQ0ZGBvz9/WFqagoAyM/Px6JFi+Dr64s6deoUuV92djY2btyIiRMnYvz48QCA9u3bo3///ti6dSsWLlxYplxhYY8wblwwHj9OEce8vVtj7VoXFmQiIiKiKkitrm4RFhYGR0dHsSADL++EI5fLcfHixWL3u3r1KtLT0xXumqOjo4O+ffsiLCysDIlkmB/SAT17BogF2dRUD/v2DUdAwFAYG+uW4dhEREREpK7UqiRHRkbC2tpaYczY2BgWFhaIjIx8434ACu1rY2ODp0+fIisrS6k80Sk1sCy0LQou0dejR2P89dcUjB7dQqnjEREREVHloFbLLVJTU2FsbFxo3MTEBCkpKUXs8d9+Ojo60NVVnNk1NjaGIAhISUmBnp5eqbLExcVBLstFkyaHIJPJULOmHjQ19TBhwq5SHYeIiIiIykdMTEy5nXOgViVZnejq6kImk6Fhw5pSRyEiIiKiImhpab31CmhKH7tcjqokY2NjpKWlFRpPSUmBiYnJG/fLyclBdna2wmxyamoqZDLZG/ctzpUrV0q9DxERERFVDWq1Jtna2rrQ2uO0tDTEx8cXWm/8+n4A8PDhQ4XxyMhI1KtXr9RLLYiIiIioelOrkty9e3eEh4cjNTVVHAsNDYWGhga6du1a7H7t2rWDkZERTpw4IY7l5ubi5MmT6N69e7lmJiIiIqKqR62WW4wZMwa7du3CtGnT4Ovri9jYWKxatQpjxoxRuEayt7c3nj59ilOnTgF4uX7Y19cXfn5+MDMzg62tLfbu3Yvk5GRMmjRJqi+HiIiIiCoptSrJJiYm2LFjB5YsWYJp06bB0NAQI0aMwMyZMxW2k8vlyM/PVxibPHkyBEHAtm3bxNtSb926VSV32yMiIiKi6kUmCAVXASYiIiIiIkDN1iQTEREREakDlmQiIiIiotewJBMRERERvYYlmYiIiIjoNSzJRERERESvYUkmIiIiInoNSzIRERER0WuqZUmOiIjAhAkT0KZNG3Tt2hWrVq1CTk7OW/cTBAGbNm1Cz5490apVK4wePRrXr18v/8BUJsp83nFxcVi1ahVcXV3Rtm1bdO/eHR9//DGio6MrKDUpS9nv71cFBATAzs4Ovr6+5ZSSVKUsn3dsbCw+/fRTdO7cGa1atYKLiwsOHz5czompLJT9vJOSkrBgwQL07NkTbdq0waBBg7B3794KSExl8ejRIyxYsACurq5o1qwZBg0aVKL9VNXX1OqOexUhJSUF3t7eaNKkCfz8/BAbG4sVK1YgKysLCxYseOO+mzdvxtq1azF79mzY2dkhMDAQEydOxKFDh3hnPzWl7Od969YtnDp1CsOHD0fr1q2RlJSE9evXY+TIkTh69CjMzMwq8KugkirL93eB+Ph4rFu3Dubm5uWclsqqLJ93XFwcRo8eDSsrKyxZsgRGRka4f/9+qX+goopTls97xowZiIyMxKxZs2BpaYmwsDAsXLgQmpqaGDVqVAV9BVRa9+/fx7lz59C6dWvI5XKU9P53KutrQjWzYcMGoU2bNkJSUpI4tm/fPuGdd94Rnj17Vux+WVlZQrt27YTVq1eLY9nZ2UKvXr2EL7/8shwTU1ko+3mnpKQIubm5CmMxMTGCnZ2dsHXr1vKKS2Wk7Of9qjlz5giffPKJ4OnpKfj4+JRTUlKFsnzes2fPFkaPHi3k5eWVc0pSFWU/77i4OMHW1lY4ePCgwriHh4fg5eVVXnFJBfLz88X//vTTT4WBAwe+dR9V9rVqt9wiLCwMjo6OMDU1FcdcXFwgl8tx8eLFYve7evUq0tPT4eLiIo7p6Oigb9++CAsLK8/IVAbKft7GxsbQ0lL8RUvdunVhZmaGuLi48opLZaTs513gypUr+OWXX/Dxxx+XY0pSFWU/7/T0dJw4cQLu7u7Q1NSsgKSkCsp+3nl5eQCAGjVqKIwbGRmVeGaSpKGhUfqaqsq+Vu1KcmRkJKytrRXGjI2NYWFhgcjIyDfuB6DQvjY2Nnj69CmysrJUH5bKTNnPuygPHz5EQkICbGxsVBmRVKgsn3d+fj6WLFmCKVOmoHbt2uUZk1RE2c/71q1byM3NhZaWFjw9PdG8eXN07doVX3/9NXJzc8s7NilJ2c/b0tIS3bp1w4YNG/DgwQOkp6fj+PHjuHjxIjw8PMo7NlUwVfa1arcmOTU1FcbGxoXGTUxMkJKS8sb9dHR0oKurqzBubGwMQRCQkpICPT09leelslH2836dIAhYunQpateujYEDB6oyIqlQWT7vPXv24MWLFxg/fnw5pSNVU/bzfv78OQDg888/x6hRo/DBBx/gxo0bWLt2LTQ0NPibBDVVlu9vPz8/zJw5U/z7W1NTE59//jmcnZ3LJStJR5V9rdqVZCJl+Pn54dKlS9iyZQsMDAykjkMqlpCQgLVr12LlypXQ0dGROg6VM7lcDgDo0qUL5s6dCwDo3LkzMjIysG3bNkybNo2THlWIIAiYN28e/v33X6xevRoWFhYIDw/HsmXLYGJiwokPKla1K8nGxsZIS0srNJ6SkgITE5M37peTk4Ps7GyFn05SU1Mhk8neuC9JR9nP+1X79+/HunXr8NVXX8HR0VHVEUmFlP28v//+e9jZ2aFDhw5ITU0F8HIdY15eHlJTU2FgYFBojTpJryx/nwMvi/GrHB0dsWHDBjx69Ah2dnaqDUtlpuznffbsWYSGhuLw4cPi5+rg4ICEhASsWLGCJbmKUWVfq3Zrkq2trQutXUpLS0N8fHyh9Suv7we8XJf6qsjISNSrV4+zDmpK2c+7wKlTp7Bw4UJMnz4dI0aMKK+YpCLKft4PHz7E5cuX0bFjR/F/V69exYULF9CxY0eEh4eXd3RSgrKfd9OmTd943OzsbJXko/9r795jorjaMIA/wKfIVUC8RJAqklmosKWoixUKiigC9QIR1KC1iEojFd3WVpDG1EsbNagt4oXYCqJmRVtXRS6CN1RQelELEaOtCbagrhUqlxZE2Pn+aNiwCyggaovPLyFx3jkz5509Yt49npnpXl0d719//RUGBgYQBEEr7uTkhPv376Ouru655EsvR3fWa69ckezl5YWCggLNbBEAZGdnQ19fHx4eHu0e5+bmBlNTU2RlZWlijx8/Rk5ODry8vJ5rztR1XR1vACgsLMSHH36IkJAQREVFPe9UqRt0dbxXrlyJ1NRUrR9HR0e4uroiNTUVUqn0RaRPndTV8baxsYEgCK2+/BQUFKBPnz5PLaLp5XiW8W5qasKNGze04teuXUO/fv1gZGT03HKmF68767VX7v8PZ82ahb179yIqKgqRkZFQqVTYuHEjZs2ahYEDB2razZs3D3fu3EFubi4AwNDQEJGRkdi6dSusrKwgCAIUCgUePnyIiIiIl3U59BRdHe9bt24hKioKQ4cOxbRp07Te1GNlZQU7O7sXfSnUAV0dbycnp1bnMjc3h7GxMdzd3V9Y/tQ5XR1vAJDL5Vi8eDE+//xzjBs3DsXFxdi9ezciIiJ438G/VFfH28vLC4MHD0Z0dDSioqIwYMAAXLhwAUqlEkuWLHlZl0MdUFdXh7y8PABAeXk5amtrkZ2dDQCQyWSwsrJ6rvXaK1ck9+3bF3v27MHatWsRFRUFExMTzJgxA3K5XKudWq1GU1OTVmzhwoUQRRG7d+9GZWUlnJyc8M033/Bte/9iXR3vn3/+GTU1NaipqcHs2bO12gYFBWH9+vUvJH/qnGf5/ab/nmcZbx8fH2zevBnbt2+HQqHAgAEDsGTJEixatOhFXgJ1QlfH29TUFCkpKdiyZQvi4+NRU1MDW1tbxMTEYM6cOS/6MqgTKioqsHTpUq1Y83Zqairc3d2fa72mJ/JJ2kREREREWl65NclERERERE/DIpmIiIiISAeLZCIiIiIiHSySiYiIiIh0sEgmIiIiItLBIpmIiIiISAeLZCIiIiIiHSySiYjwz2vIJRIJCgsLX3Yqz5VEIsHWrVs71NbHxwcxMTHPOSMion+nV+6Ne0TUsxw+fBixsbFt7lu4cCGWL1/+gjPqON3ce/fujcGDB8PDwwOLFy+GtbX1c8/h8uXLyM/Px7x582Bubv7c++sIHx8flJeXa7aNjIzg4OCAOXPmYPr06V06Z15eHoqKivgaYiLqMBbJRNQjREdHw9bWVismCMJLyqZzmnNvaGjATz/9BIVCgby8PBw/fhxGRkbd2ldRUREMDAw021euXEFiYiKCgoJaFcnZ2dnQ09Pr1v47ysnJCeHh4QCAP/74A4cOHcKKFSvQ0NCA0NDQTp8vLy8P+/fvZ5FMRB3GIpmIegQvLy+4uLi87DS6pGXuISEhsLCwQHJyMk6dOoV33nmnW/syNDTscNvevXt3a9+dMXDgQEybNk2zHRwcjAkTJiAlJaVLRTIRUWdxTTIR9Wjl5eX47LPP4OfnB6lUCnd3d0RHR6OsrOypx5aWlmLJkiXw8PCAi4sLvLy8IJfLUVNTo9Xu6NGjCA4OhlQqhUwmg1wux927d7uc85gxYwBAk2NjYyO2bdsGX19fODs7w8fHB5s3b0ZDQ4PWccXFxYiIiIC7uzukUil8fHxaLUVpuSZ569at2LhxIwBgwoQJkEgkkEgkmn5brkkuLi6GRCKBUqlsle/58+chkUhw5swZTUylUiE2NhZjx46Fs7MzAgMD8e2333b5M7GysoK9vT1+++03rfiPP/6I6OhojBs3Ds7OzvD29sYXX3yB+vp6TZuYmBjs379fc/3NP83UajVSUlIQGBgIFxcXjB07FqtWrUJVVVWX8yWi/z7OJBNRj1BbW4vKykqtmJWVFYqLi3HlyhUEBgZi0KBBKC8vh0KhwLvvvouMjIx2lzM0NDQgIiICDQ0NmDNnDqytraFSqXD27FlUV1fDzMwMALBjxw589dVX8Pf3x4wZM1BZWYl9+/YhLCwMR44c6dI63+ZC0MLCAgDw6aefQqlUws/PD+Hh4SgqKkJSUhJu3bqFbdu2AQAqKioQEREBS0tLLFq0CObm5igrK0Nubm67/UycOBGlpaU4fvw4YmNjYWlpqfncdLm4uGDIkCHIyspCUFCQ1r7MzEz07dsXnp6eAIAHDx4gNDQUenp6CAsLg5WVFc6dO4e4uDjU1tbivffe6/Rn0tjYCJVKhb59+2rFs7OzUV9fj9mzZ8PCwgJFRUXYt28f7t27h4SEBADAzJkzcf/+feTn52u+FLS0atUqKJVKBAcHY+7cuSgrK8P+/ftRUlIChUKBXr16dTpfIuoBRCKi/7DvvvtOFAShzR9RFMW6urpWx1y5ckUUBEFUKpWa2KVLl0RBEMRLly6JoiiKJSUloiAIYlZWVrt9l5WViU5OTuKOHTu04jdu3BBff/31VvH2ci8oKBArKirEu3fvihkZGaJMJhOlUql479498fr166IgCGJcXJzWsevXrxcFQRAvXrwoiqIo5ubmioIgiEVFRU/sUxAEMSEhQbP99ddfi4IgiL///nurtuPHjxdXrFih2d60aZM4YsQI8eHDh5rYo0ePxFGjRomxsbGa2MqVK0UPDw+xsrJS63xyuVwcOXJkm2Oi2+/8+fPFiooKsaKiQrxx44b48ccfi4IgiKtXr9Zq29a5kpKSRIlEIpaXl2tiq1ev1vydaOmHH34QBUEQjx07phU/d+5cm3EienVwJpmIeoRVq1Zh2LBhreJ9+vTR/Pnx48eora2FnZ0dzM3NUVJS0u7TEkxNTQEAFy5cgLe3d5szzrm5uVCr1fD399eaxba2tsZrr72GwsJCvP/++0/NXXdm1cbGBvHx8Rg4cCCOHDkCAJqb2JrNnz8fu3fvRl5eHsaMGaOZ2T579iwcHR2fy+xnQEAAkpKSkJOTg5CQEABAfn4+qqurERAQAAAQRRE5OTnw9/eHKIpan4unpycyMjJw7do1jBw58ol9XbhwAW+99ZZWLDg4GJ988olWrOX4/v3336ivr8ebb74JURRRUlKCwYMHP7Gf7OxsmJmZwcPDQyvXESNGwNjYGIWFhZgyZcoTz0FEPROLZCLqEaRSaZs37tXX1yMpKQmHDx+GSqWCKIqafbpri1saMmQIwsPDkZycjPT0dIwaNQo+Pj6YOnWqpiAtLS2FKIqYNGlSm+f43/869k9sc4FvYGAAa2trDBs2DPr6/9wyUl5eDn19fdjZ2Wkd079/f5ibm2selSaTyeDn54fExESkpKRAJpPB19cXU6ZM6bYb8BwdHWFvb4+srCxNkZyZmQlLS0vNOurKykpUV1cjLS0NaWlpbZ5Hd1lMW9544w0sW7YMTU1N+OWXX7Bjxw5UV1e3Kv7v3LmDhIQEnD59utUa4tra2qf2c/v2bdTU1LQqyJtVVFQ89RxE1DOxSCaiHm3t2rU4fPgw5s2bB1dXV5iZmUFPTw9yuVyrYG5LTEwMgoKCcOrUKeTn52PdunVISkrCwYMHMWjQIKjVaujp6WHXrl1aj1VrZmxs3KEc2yvwW3rao9j09PSQkJCAq1ev4syZMzh//jxWrlyJ5ORkpKWlwcTEpEO5PE1AQAB27tyJyspKmJqa4vTp0wgMDNR8IVCr1QCAqVOntlq73KzlTXPtsbS0xNixYwEAb7/9Nuzt7REZGYnU1FTNrHpTUxPCw8NRVVWFBQsWwN7eHsbGxlCpVIiJidHk8iRqtRr9+vVDfHx8m/vbWp9NRK8GFslE1KOdOHEC06dP13pz3KNHj544i9xS85MQFi9ejMuXL2P27NlQKBSQy+Wws7ODKIqwtbVtc6lHd7CxsYFarcbt27cxfPhwTfzBgweorq6GjY2NVntXV1e4urpCLpcjPT0dy5cvR2ZmpmbmV1dnn4McEBCAxMRE5OTkwNraGrW1tQgMDNTst7KygomJCdRqtabI7Q7jxo2DTCbDzp07MXPmTBgbG+PmzZsoLS3Fhg0btJbN5Ofntzq+veu0s7PDxYsX4ebmprV0g4iIj4Ajoh6trRnevXv3oqmp6YnH1dbWorGxUSsmCAL09fU1j16bNGkSDAwMkJiY2GpWWhRF/Pnnn8+YPeDt7Q0A2LNnj1Y8OTlZa39VVVWrHJycnACg1aPiWmpea93RLw3Dhw+HIAjIzMxEZmYm+vfvj9GjR2v2GxgYwM/PDydOnMDNmzdbHd+RpRbtWbBgAR4+fIiDBw8CgGZJSsvrFkURqamprY5tvs7q6mqtuL+/P5qamrB9+/ZWxzQ2NrZqT0SvDs4kE1GPNm7cOBw9ehSmpqZwcHDA1atXUVBQoHm8WnsuXbqENWvWYPLkyRg6dCiamppw9OhRTREI/DMLuWzZMmzatAnl5eXw9fWFiYkJysrKcPLkSYSGhiIiIuKZ8nd0dERQUBDS0tJQXV2N0aNHo7i4GEqlEr6+vpq1wEqlEgqFAr6+vrCzs8Nff/2FgwcPwtTUFF5eXu2ef8SIEQCALVu2ICAgAL169cL48eOfuFQkICAACQkJMDQ0xIwZMzTFarOPPvoIhYWFCA0NRUhICBwcHFBVVYVr167h4sWL+P7777v0WXh7e0MQBKSkpCAsLAz29vaws7PDhg0boFKpYGpqihMnTrRZ2DZf57p16+Dp6QkDAwMEBgZCJpNh5syZSEpKwvXr1+Hh4YFevXqhtLQU2dnZiIuLw+TJk7uULxH9t7FIJqIeLS4uDvr6+khPT8ejR4/g5uaG5ORkLFiw4InHSSQSeHp64syZM1CpVDAyMoJEIsGuXbvg6uqqabdo0SIMHToUKSkpmmcWDxo0CB4eHvDx8emWa1i3bh1sbW2hVCpx8uRJWFtbIzIyEh988IGmjUwmQ3FxMTIzM/HgwQOYmZlBKpUiPj4eQ4YMaffcUqkUS5cuxYEDB3D+/Hmo1WqcOnXqqUXyl19+ibq6Ovj7+7fab21tjUOHDmHbtm3Izc2FQqGAhYUFHBwcsHz58mf6LObPn4+YmBikp6cjODgYO3fu1KwVNzQ0xMSJExEWFqb1tj7gn1n/uXPnIiMjA8eOHYMoipplImvWrIGzszMOHDiALVu2wMDAADY2Npg6dSrc3NyeKV8i+u/SE5925woRERER0SuGa5KJiIiIiHSwSCYiIiIi0sEimYiIiIhIB4tkIiIiIiIdLJKJiIiIiHSwSCYiIiIi0sEimYiIiIhIB4tkIiIiIiIdLJKJiIiIiHSwSCYiIiIi0sEimYiIiIhIB4tkIiIiIiIdLJKJiIiIiHT8H3vRv+m2vwOOAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Calculate the ROC curve\n",
    "fpr, tpr, thresholds = roc_curve(y_test, y_prediction)\n",
    "\n",
    "# Calculate the area under the ROC curve (AUC)\n",
    "roc_auc = auc(fpr, tpr)\n",
    "\n",
    "# Plot the ROC curve\n",
    "plt.figure(figsize=(8,6), dpi = 100)\n",
    "plt.plot(fpr, tpr, color='darkorange', lw=2, label='ROC curve (area = %0.2f)' % roc_auc)\n",
    "plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver Operating Characteristic (ROC)')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.savefig(\"ROC_LSTM_2_feature_selection_500_epoches.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "45e3332c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "      normal       0.99      1.00      0.99      1056\n",
      "     anomaly       0.91      0.43      0.59        23\n",
      "\n",
      "    accuracy                           0.99      1079\n",
      "   macro avg       0.95      0.72      0.79      1079\n",
      "weighted avg       0.99      0.99      0.98      1079\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Print classification report\n",
    "print(classification_report(y_test, y_predict, target_names=['normal', 'anomaly']))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe9786dc",
   "metadata": {},
   "source": [
    "# Confustion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "11823e22",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "\n",
    "# Define class labels\n",
    "class_labels = ['Normal', 'Anomaly']\n",
    "\n",
    "conf_matrix = confusion_matrix(y_test, y_predict)\n",
    "\n",
    "\n",
    "# Set up a white grid with white text\n",
    "sns.set(style=\"white\")\n",
    "plt.figure(figsize=(6, 4), dpi=100)\n",
    "\n",
    "# Create a heatmap using Seaborn\n",
    "sns.heatmap(conf_matrix, annot=True, fmt=\"d\", cmap=\"YlGnBu\", cbar=False,\n",
    "            annot_kws={\"fontsize\": 12}, linewidths=1, linecolor='black')\n",
    "\n",
    "# Set custom tick labels for x-axis and y-axis\n",
    "tick_labels = class_labels\n",
    "plt.xticks(np.arange(len(tick_labels)) + 0.5, tick_labels, fontsize=12)\n",
    "plt.yticks(np.arange(len(tick_labels)) + 0.5, tick_labels, fontsize=12)\n",
    "\n",
    "# Add labels and title with increased font size\n",
    "plt.xlabel('Predicted', fontsize=12)\n",
    "plt.ylabel('Actual', fontsize=12)\n",
    "plt.title('Confusion Matrix', fontsize=14)\n",
    "\n",
    "# Set tick labels' font size\n",
    "plt.xticks(fontsize=12)\n",
    "plt.yticks(fontsize=12)\n",
    "\n",
    "plt.savefig(\"Confusion_matrix_lstm_2_feature_selection_500_epoches.png\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3f750f87",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "43337e83",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "536d6535",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.17"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}