NN-Loan_Approval.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "7b9c40b1-de47-4f89-b5ae-9b0521481c91",
   "metadata": {},
   "source": [
    "# Loan Approval / Modeling"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aba84977-1fba-49bc-b34a-964092042a42",
   "metadata": {},
   "source": [
    "## Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "8e09dc64-51fe-4201-a33b-6f03f4fd4baf",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import pickle\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay, f1_score\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "3be13d06-1866-4803-843a-99ebbd7dc298",
   "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>Loan_ID</th>\n",
       "      <th>ApplicantIncome</th>\n",
       "      <th>CoapplicantIncome</th>\n",
       "      <th>LoanAmount</th>\n",
       "      <th>Loan_Amount_Term</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Loan_Status</th>\n",
       "      <th>Gender_Female</th>\n",
       "      <th>Gender_Male</th>\n",
       "      <th>Married_No</th>\n",
       "      <th>...</th>\n",
       "      <th>Dependents_1</th>\n",
       "      <th>Dependents_2</th>\n",
       "      <th>Dependents_3+</th>\n",
       "      <th>Education_Graduate</th>\n",
       "      <th>Education_Not Graduate</th>\n",
       "      <th>Self_Employed_No</th>\n",
       "      <th>Self_Employed_Yes</th>\n",
       "      <th>Property_Area_Rural</th>\n",
       "      <th>Property_Area_Semiurban</th>\n",
       "      <th>Property_Area_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LP001002</td>\n",
       "      <td>0.070489</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.172214</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LP001003</td>\n",
       "      <td>0.054830</td>\n",
       "      <td>0.036192</td>\n",
       "      <td>0.172214</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LP001005</td>\n",
       "      <td>0.035250</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.082489</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>LP001006</td>\n",
       "      <td>0.030093</td>\n",
       "      <td>0.056592</td>\n",
       "      <td>0.160637</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LP001008</td>\n",
       "      <td>0.072356</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.191027</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    Loan_ID  ApplicantIncome  CoapplicantIncome  LoanAmount  Loan_Amount_Term  \\\n",
       "0  LP001002         0.070489           0.000000    0.172214           0.74359   \n",
       "1  LP001003         0.054830           0.036192    0.172214           0.74359   \n",
       "2  LP001005         0.035250           0.000000    0.082489           0.74359   \n",
       "3  LP001006         0.030093           0.056592    0.160637           0.74359   \n",
       "4  LP001008         0.072356           0.000000    0.191027           0.74359   \n",
       "\n",
       "   Credit_History  Loan_Status  Gender_Female  Gender_Male  Married_No  ...  \\\n",
       "0             1.0            1            0.0          1.0         1.0  ...   \n",
       "1             1.0            0            0.0          1.0         0.0  ...   \n",
       "2             1.0            1            0.0          1.0         0.0  ...   \n",
       "3             1.0            1            0.0          1.0         0.0  ...   \n",
       "4             1.0            1            0.0          1.0         1.0  ...   \n",
       "\n",
       "   Dependents_1  Dependents_2  Dependents_3+  Education_Graduate  \\\n",
       "0           0.0           0.0            0.0                 1.0   \n",
       "1           1.0           0.0            0.0                 1.0   \n",
       "2           0.0           0.0            0.0                 1.0   \n",
       "3           0.0           0.0            0.0                 0.0   \n",
       "4           0.0           0.0            0.0                 1.0   \n",
       "\n",
       "   Education_Not Graduate  Self_Employed_No  Self_Employed_Yes  \\\n",
       "0                     0.0               1.0                0.0   \n",
       "1                     0.0               1.0                0.0   \n",
       "2                     0.0               0.0                1.0   \n",
       "3                     1.0               1.0                0.0   \n",
       "4                     0.0               1.0                0.0   \n",
       "\n",
       "   Property_Area_Rural  Property_Area_Semiurban  Property_Area_Urban  \n",
       "0                  0.0                      0.0                  1.0  \n",
       "1                  1.0                      0.0                  0.0  \n",
       "2                  0.0                      0.0                  1.0  \n",
       "3                  0.0                      0.0                  1.0  \n",
       "4                  0.0                      0.0                  1.0  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"../data/loan-processed.csv\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "80330147-5b0e-49fb-b85d-cf5077856969",
   "metadata": {},
   "source": [
    "## Feature Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "176071fd-5233-4b64-911c-fec2cf3e9a11",
   "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>ApplicantIncome</th>\n",
       "      <th>CoapplicantIncome</th>\n",
       "      <th>LoanAmount</th>\n",
       "      <th>Loan_Amount_Term</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Loan_Status</th>\n",
       "      <th>Gender_Female</th>\n",
       "      <th>Gender_Male</th>\n",
       "      <th>Married_No</th>\n",
       "      <th>Married_Yes</th>\n",
       "      <th>...</th>\n",
       "      <th>Dependents_1</th>\n",
       "      <th>Dependents_2</th>\n",
       "      <th>Dependents_3+</th>\n",
       "      <th>Education_Graduate</th>\n",
       "      <th>Education_Not Graduate</th>\n",
       "      <th>Self_Employed_No</th>\n",
       "      <th>Self_Employed_Yes</th>\n",
       "      <th>Property_Area_Rural</th>\n",
       "      <th>Property_Area_Semiurban</th>\n",
       "      <th>Property_Area_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.070489</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.172214</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.054830</td>\n",
       "      <td>0.036192</td>\n",
       "      <td>0.172214</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.035250</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.082489</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.030093</td>\n",
       "      <td>0.056592</td>\n",
       "      <td>0.160637</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.072356</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.191027</td>\n",
       "      <td>0.74359</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ApplicantIncome  CoapplicantIncome  LoanAmount  Loan_Amount_Term  \\\n",
       "0         0.070489           0.000000    0.172214           0.74359   \n",
       "1         0.054830           0.036192    0.172214           0.74359   \n",
       "2         0.035250           0.000000    0.082489           0.74359   \n",
       "3         0.030093           0.056592    0.160637           0.74359   \n",
       "4         0.072356           0.000000    0.191027           0.74359   \n",
       "\n",
       "   Credit_History  Loan_Status  Gender_Female  Gender_Male  Married_No  \\\n",
       "0             1.0            1            0.0          1.0         1.0   \n",
       "1             1.0            0            0.0          1.0         0.0   \n",
       "2             1.0            1            0.0          1.0         0.0   \n",
       "3             1.0            1            0.0          1.0         0.0   \n",
       "4             1.0            1            0.0          1.0         1.0   \n",
       "\n",
       "   Married_Yes  ...  Dependents_1  Dependents_2  Dependents_3+  \\\n",
       "0          0.0  ...           0.0           0.0            0.0   \n",
       "1          1.0  ...           1.0           0.0            0.0   \n",
       "2          1.0  ...           0.0           0.0            0.0   \n",
       "3          1.0  ...           0.0           0.0            0.0   \n",
       "4          0.0  ...           0.0           0.0            0.0   \n",
       "\n",
       "   Education_Graduate  Education_Not Graduate  Self_Employed_No  \\\n",
       "0                 1.0                     0.0               1.0   \n",
       "1                 1.0                     0.0               1.0   \n",
       "2                 1.0                     0.0               0.0   \n",
       "3                 0.0                     1.0               1.0   \n",
       "4                 1.0                     0.0               1.0   \n",
       "\n",
       "   Self_Employed_Yes  Property_Area_Rural  Property_Area_Semiurban  \\\n",
       "0                0.0                  0.0                      0.0   \n",
       "1                0.0                  1.0                      0.0   \n",
       "2                1.0                  0.0                      0.0   \n",
       "3                0.0                  0.0                      0.0   \n",
       "4                0.0                  0.0                      0.0   \n",
       "\n",
       "   Property_Area_Urban  \n",
       "0                  1.0  \n",
       "1                  0.0  \n",
       "2                  1.0  \n",
       "3                  1.0  \n",
       "4                  1.0  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.drop(columns=[\"Loan_ID\"], inplace=True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "acdfb87b-170c-4875-8691-f35c462f503f",
   "metadata": {},
   "source": [
    "## Train / Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "c0160978-3191-4625-b1c2-a4543ebeb4c1",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df.drop(columns=[\"Loan_Status\"])\n",
    "y = df[\"Loan_Status\"]\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.20, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "73a7645e-bec5-47bb-af08-a1bf508346f9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(491, 20)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "0b4b545b-c5ce-4ffb-9075-29333bf16938",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(123, 20)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8be43640-76ad-4f68-bf7d-5721ec962023",
   "metadata": {},
   "source": [
    "## Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "fc723043-b9b2-4f0d-9600-c67f6d110038",
   "metadata": {
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iteration 1, loss = 0.67810580\n",
      "Validation score: 0.620000\n",
      "Iteration 2, loss = 0.67388725\n",
      "Validation score: 0.620000\n",
      "Iteration 3, loss = 0.66984980\n",
      "Validation score: 0.620000\n",
      "Iteration 4, loss = 0.66625912\n",
      "Validation score: 0.620000\n",
      "Iteration 5, loss = 0.66303761\n",
      "Validation score: 0.640000\n",
      "Iteration 6, loss = 0.65989442\n",
      "Validation score: 0.660000\n",
      "Iteration 7, loss = 0.65736001\n",
      "Validation score: 0.660000\n",
      "Iteration 8, loss = 0.65482643\n",
      "Validation score: 0.660000\n",
      "Iteration 9, loss = 0.65247698\n",
      "Validation score: 0.660000\n",
      "Iteration 10, loss = 0.65036226\n",
      "Validation score: 0.680000\n",
      "Iteration 11, loss = 0.64851318\n",
      "Validation score: 0.680000\n",
      "Iteration 12, loss = 0.64658474\n",
      "Validation score: 0.680000\n",
      "Iteration 13, loss = 0.64486524\n",
      "Validation score: 0.680000\n",
      "Iteration 14, loss = 0.64325677\n",
      "Validation score: 0.680000\n",
      "Iteration 15, loss = 0.64181771\n",
      "Validation score: 0.680000\n",
      "Iteration 16, loss = 0.64039440\n",
      "Validation score: 0.680000\n",
      "Iteration 17, loss = 0.63901162\n",
      "Validation score: 0.680000\n",
      "Iteration 18, loss = 0.63773269\n",
      "Validation score: 0.680000\n",
      "Iteration 19, loss = 0.63618377\n",
      "Validation score: 0.680000\n",
      "Iteration 20, loss = 0.63468684\n",
      "Validation score: 0.680000\n",
      "Iteration 21, loss = 0.63312991\n",
      "Validation score: 0.680000\n",
      "Iteration 22, loss = 0.63173265\n",
      "Validation score: 0.680000\n",
      "Iteration 23, loss = 0.63020280\n",
      "Validation score: 0.680000\n",
      "Iteration 24, loss = 0.62859769\n",
      "Validation score: 0.700000\n",
      "Iteration 25, loss = 0.62727468\n",
      "Validation score: 0.700000\n",
      "Iteration 26, loss = 0.62593863\n",
      "Validation score: 0.700000\n",
      "Iteration 27, loss = 0.62472873\n",
      "Validation score: 0.700000\n",
      "Iteration 28, loss = 0.62358139\n",
      "Validation score: 0.700000\n",
      "Iteration 29, loss = 0.62237810\n",
      "Validation score: 0.700000\n",
      "Iteration 30, loss = 0.62125300\n",
      "Validation score: 0.700000\n",
      "Iteration 31, loss = 0.62016945\n",
      "Validation score: 0.700000\n",
      "Iteration 32, loss = 0.61905797\n",
      "Validation score: 0.700000\n",
      "Iteration 33, loss = 0.61800153\n",
      "Validation score: 0.700000\n",
      "Iteration 34, loss = 0.61691789\n",
      "Validation score: 0.700000\n",
      "Iteration 35, loss = 0.61587562\n",
      "Validation score: 0.700000\n",
      "Iteration 36, loss = 0.61488978\n",
      "Validation score: 0.700000\n",
      "Iteration 37, loss = 0.61385307\n",
      "Validation score: 0.700000\n",
      "Iteration 38, loss = 0.61277800\n",
      "Validation score: 0.700000\n",
      "Iteration 39, loss = 0.61177399\n",
      "Validation score: 0.700000\n",
      "Iteration 40, loss = 0.61082725\n",
      "Validation score: 0.700000\n",
      "Iteration 41, loss = 0.60974934\n",
      "Validation score: 0.700000\n",
      "Iteration 42, loss = 0.60883824\n",
      "Validation score: 0.700000\n",
      "Iteration 43, loss = 0.60791870\n",
      "Validation score: 0.700000\n",
      "Iteration 44, loss = 0.60704243\n",
      "Validation score: 0.700000\n",
      "Iteration 45, loss = 0.60608155\n",
      "Validation score: 0.700000\n",
      "Iteration 46, loss = 0.60521269\n",
      "Validation score: 0.700000\n",
      "Iteration 47, loss = 0.60433738\n",
      "Validation score: 0.700000\n",
      "Iteration 48, loss = 0.60341527\n",
      "Validation score: 0.700000\n",
      "Iteration 49, loss = 0.60253677\n",
      "Validation score: 0.700000\n",
      "Iteration 50, loss = 0.60157057\n",
      "Validation score: 0.700000\n",
      "Iteration 51, loss = 0.60063623\n",
      "Validation score: 0.700000\n",
      "Iteration 52, loss = 0.59965619\n",
      "Validation score: 0.700000\n",
      "Iteration 53, loss = 0.59875487\n",
      "Validation score: 0.700000\n",
      "Iteration 54, loss = 0.59784747\n",
      "Validation score: 0.700000\n",
      "Iteration 55, loss = 0.59685710\n",
      "Validation score: 0.700000\n",
      "Iteration 56, loss = 0.59598673\n",
      "Validation score: 0.700000\n",
      "Iteration 57, loss = 0.59499133\n",
      "Validation score: 0.700000\n",
      "Iteration 58, loss = 0.59410710\n",
      "Validation score: 0.700000\n",
      "Iteration 59, loss = 0.59313649\n",
      "Validation score: 0.700000\n",
      "Iteration 60, loss = 0.59227101\n",
      "Validation score: 0.700000\n",
      "Iteration 61, loss = 0.59135705\n",
      "Validation score: 0.700000\n",
      "Iteration 62, loss = 0.59053214\n",
      "Validation score: 0.700000\n",
      "Iteration 63, loss = 0.58959846\n",
      "Validation score: 0.700000\n",
      "Iteration 64, loss = 0.58859341\n",
      "Validation score: 0.720000\n",
      "Iteration 65, loss = 0.58752137\n",
      "Validation score: 0.720000\n",
      "Iteration 66, loss = 0.58658466\n",
      "Validation score: 0.720000\n",
      "Iteration 67, loss = 0.58568271\n",
      "Validation score: 0.720000\n",
      "Iteration 68, loss = 0.58478202\n",
      "Validation score: 0.720000\n",
      "Iteration 69, loss = 0.58395444\n",
      "Validation score: 0.720000\n",
      "Iteration 70, loss = 0.58310291\n",
      "Validation score: 0.720000\n",
      "Iteration 71, loss = 0.58216867\n",
      "Validation score: 0.720000\n",
      "Iteration 72, loss = 0.58112899\n",
      "Validation score: 0.720000\n",
      "Iteration 73, loss = 0.58005797\n",
      "Validation score: 0.720000\n",
      "Iteration 74, loss = 0.57895577\n",
      "Validation score: 0.720000\n",
      "Iteration 75, loss = 0.57799787\n",
      "Validation score: 0.720000\n",
      "Iteration 76, loss = 0.57702822\n",
      "Validation score: 0.720000\n",
      "Iteration 77, loss = 0.57612389\n",
      "Validation score: 0.720000\n",
      "Iteration 78, loss = 0.57516546\n",
      "Validation score: 0.720000\n",
      "Iteration 79, loss = 0.57424296\n",
      "Validation score: 0.720000\n",
      "Iteration 80, loss = 0.57319833\n",
      "Validation score: 0.720000\n",
      "Iteration 81, loss = 0.57208029\n",
      "Validation score: 0.720000\n",
      "Iteration 82, loss = 0.57121432\n",
      "Validation score: 0.720000\n",
      "Iteration 83, loss = 0.57007628\n",
      "Validation score: 0.720000\n",
      "Iteration 84, loss = 0.56918727\n",
      "Validation score: 0.720000\n",
      "Iteration 85, loss = 0.56815280\n",
      "Validation score: 0.720000\n",
      "Iteration 86, loss = 0.56713298\n",
      "Validation score: 0.720000\n",
      "Iteration 87, loss = 0.56620127\n",
      "Validation score: 0.720000\n",
      "Iteration 88, loss = 0.56525859\n",
      "Validation score: 0.720000\n",
      "Iteration 89, loss = 0.56432215\n",
      "Validation score: 0.720000\n",
      "Iteration 90, loss = 0.56338801\n",
      "Validation score: 0.720000\n",
      "Iteration 91, loss = 0.56246529\n",
      "Validation score: 0.720000\n",
      "Iteration 92, loss = 0.56153328\n",
      "Validation score: 0.720000\n",
      "Iteration 93, loss = 0.56054068\n",
      "Validation score: 0.720000\n",
      "Iteration 94, loss = 0.55968449\n",
      "Validation score: 0.720000\n",
      "Iteration 95, loss = 0.55878270\n",
      "Validation score: 0.720000\n",
      "Iteration 96, loss = 0.55791819\n",
      "Validation score: 0.720000\n",
      "Iteration 97, loss = 0.55698084\n",
      "Validation score: 0.720000\n",
      "Iteration 98, loss = 0.55626152\n",
      "Validation score: 0.720000\n",
      "Iteration 99, loss = 0.55508713\n",
      "Validation score: 0.720000\n",
      "Iteration 100, loss = 0.55431817\n",
      "Validation score: 0.720000\n",
      "Iteration 101, loss = 0.55351305\n",
      "Validation score: 0.720000\n",
      "Iteration 102, loss = 0.55259473\n",
      "Validation score: 0.720000\n",
      "Iteration 103, loss = 0.55178485\n",
      "Validation score: 0.720000\n",
      "Iteration 104, loss = 0.55100701\n",
      "Validation score: 0.740000\n",
      "Iteration 105, loss = 0.55016617\n",
      "Validation score: 0.740000\n",
      "Iteration 106, loss = 0.54934943\n",
      "Validation score: 0.740000\n",
      "Iteration 107, loss = 0.54844992\n",
      "Validation score: 0.740000\n",
      "Iteration 108, loss = 0.54763031\n",
      "Validation score: 0.740000\n",
      "Iteration 109, loss = 0.54669418\n",
      "Validation score: 0.740000\n",
      "Iteration 110, loss = 0.54586315\n",
      "Validation score: 0.740000\n",
      "Iteration 111, loss = 0.54508277\n",
      "Validation score: 0.760000\n",
      "Iteration 112, loss = 0.54413234\n",
      "Validation score: 0.760000\n",
      "Iteration 113, loss = 0.54321253\n",
      "Validation score: 0.740000\n",
      "Iteration 114, loss = 0.54242665\n",
      "Validation score: 0.740000\n",
      "Iteration 115, loss = 0.54157916\n",
      "Validation score: 0.740000\n",
      "Iteration 116, loss = 0.54080231\n",
      "Validation score: 0.740000\n",
      "Iteration 117, loss = 0.53995920\n",
      "Validation score: 0.740000\n",
      "Iteration 118, loss = 0.53909661\n",
      "Validation score: 0.740000\n",
      "Iteration 119, loss = 0.53834134\n",
      "Validation score: 0.740000\n",
      "Iteration 120, loss = 0.53760500\n",
      "Validation score: 0.740000\n",
      "Iteration 121, loss = 0.53692827\n",
      "Validation score: 0.740000\n",
      "Iteration 122, loss = 0.53620369\n",
      "Validation score: 0.740000\n",
      "Iteration 123, loss = 0.53547336\n",
      "Validation score: 0.740000\n",
      "Iteration 124, loss = 0.53474887\n",
      "Validation score: 0.740000\n",
      "Iteration 125, loss = 0.53406552\n",
      "Validation score: 0.740000\n",
      "Iteration 126, loss = 0.53343762\n",
      "Validation score: 0.740000\n",
      "Iteration 127, loss = 0.53276090\n",
      "Validation score: 0.740000\n",
      "Iteration 128, loss = 0.53198550\n",
      "Validation score: 0.740000\n",
      "Iteration 129, loss = 0.53128994\n",
      "Validation score: 0.740000\n",
      "Iteration 130, loss = 0.53047405\n",
      "Validation score: 0.740000\n",
      "Iteration 131, loss = 0.52975368\n",
      "Validation score: 0.740000\n",
      "Iteration 132, loss = 0.52894584\n",
      "Validation score: 0.740000\n",
      "Iteration 133, loss = 0.52814104\n",
      "Validation score: 0.740000\n",
      "Iteration 134, loss = 0.52737781\n",
      "Validation score: 0.740000\n",
      "Iteration 135, loss = 0.52664533\n",
      "Validation score: 0.740000\n",
      "Iteration 136, loss = 0.52581342\n",
      "Validation score: 0.740000\n",
      "Iteration 137, loss = 0.52502630\n",
      "Validation score: 0.740000\n",
      "Iteration 138, loss = 0.52436927\n",
      "Validation score: 0.740000\n",
      "Iteration 139, loss = 0.52361189\n",
      "Validation score: 0.740000\n",
      "Iteration 140, loss = 0.52293864\n",
      "Validation score: 0.740000\n",
      "Iteration 141, loss = 0.52227138\n",
      "Validation score: 0.740000\n",
      "Iteration 142, loss = 0.52154335\n",
      "Validation score: 0.740000\n",
      "Iteration 143, loss = 0.52073004\n",
      "Validation score: 0.740000\n",
      "Iteration 144, loss = 0.52005080\n",
      "Validation score: 0.740000\n",
      "Iteration 145, loss = 0.51923652\n",
      "Validation score: 0.740000\n",
      "Iteration 146, loss = 0.51844222\n",
      "Validation score: 0.740000\n",
      "Iteration 147, loss = 0.51775396\n",
      "Validation score: 0.740000\n",
      "Iteration 148, loss = 0.51689422\n",
      "Validation score: 0.740000\n",
      "Iteration 149, loss = 0.51607911\n",
      "Validation score: 0.740000\n",
      "Iteration 150, loss = 0.51527304\n",
      "Validation score: 0.740000\n",
      "Iteration 151, loss = 0.51457421\n",
      "Validation score: 0.740000\n",
      "Iteration 152, loss = 0.51371449\n",
      "Validation score: 0.740000\n",
      "Iteration 153, loss = 0.51305558\n",
      "Validation score: 0.740000\n",
      "Iteration 154, loss = 0.51222279\n",
      "Validation score: 0.740000\n",
      "Iteration 155, loss = 0.51141479\n",
      "Validation score: 0.740000\n",
      "Iteration 156, loss = 0.51076174\n",
      "Validation score: 0.740000\n",
      "Iteration 157, loss = 0.51003026\n",
      "Validation score: 0.740000\n",
      "Iteration 158, loss = 0.50936868\n",
      "Validation score: 0.740000\n",
      "Iteration 159, loss = 0.50845933\n",
      "Validation score: 0.740000\n",
      "Iteration 160, loss = 0.50768847\n",
      "Validation score: 0.740000\n",
      "Iteration 161, loss = 0.50717101\n",
      "Validation score: 0.740000\n",
      "Iteration 162, loss = 0.50636834\n",
      "Validation score: 0.740000\n",
      "Iteration 163, loss = 0.50578151\n",
      "Validation score: 0.740000\n",
      "Iteration 164, loss = 0.50501169\n",
      "Validation score: 0.740000\n",
      "Iteration 165, loss = 0.50425187\n",
      "Validation score: 0.740000\n",
      "Iteration 166, loss = 0.50344250\n",
      "Validation score: 0.740000\n",
      "Iteration 167, loss = 0.50278967\n",
      "Validation score: 0.740000\n",
      "Iteration 168, loss = 0.50219826\n",
      "Validation score: 0.740000\n",
      "Iteration 169, loss = 0.50141607\n",
      "Validation score: 0.740000\n",
      "Iteration 170, loss = 0.50063605\n",
      "Validation score: 0.740000\n",
      "Iteration 171, loss = 0.49999626\n",
      "Validation score: 0.740000\n",
      "Iteration 172, loss = 0.49945051\n",
      "Validation score: 0.740000\n",
      "Iteration 173, loss = 0.49889272\n",
      "Validation score: 0.740000\n",
      "Iteration 174, loss = 0.49833647\n",
      "Validation score: 0.740000\n",
      "Iteration 175, loss = 0.49767357\n",
      "Validation score: 0.740000\n",
      "Iteration 176, loss = 0.49698467\n",
      "Validation score: 0.740000\n",
      "Iteration 177, loss = 0.49637643\n",
      "Validation score: 0.740000\n",
      "Iteration 178, loss = 0.49580635\n",
      "Validation score: 0.740000\n",
      "Iteration 179, loss = 0.49524445\n",
      "Validation score: 0.740000\n",
      "Iteration 180, loss = 0.49450959\n",
      "Validation score: 0.740000\n",
      "Iteration 181, loss = 0.49384026\n",
      "Validation score: 0.740000\n",
      "Iteration 182, loss = 0.49313018\n",
      "Validation score: 0.740000\n",
      "Iteration 183, loss = 0.49237364\n",
      "Validation score: 0.740000\n",
      "Iteration 184, loss = 0.49164546\n",
      "Validation score: 0.740000\n",
      "Iteration 185, loss = 0.49092203\n",
      "Validation score: 0.740000\n",
      "Iteration 186, loss = 0.49024091\n",
      "Validation score: 0.740000\n",
      "Iteration 187, loss = 0.48948451\n",
      "Validation score: 0.740000\n",
      "Iteration 188, loss = 0.48879920\n",
      "Validation score: 0.740000\n",
      "Iteration 189, loss = 0.48820087\n",
      "Validation score: 0.740000\n",
      "Iteration 190, loss = 0.48757068\n",
      "Validation score: 0.740000\n",
      "Iteration 191, loss = 0.48691299\n",
      "Validation score: 0.740000\n",
      "Iteration 192, loss = 0.48624448\n",
      "Validation score: 0.740000\n",
      "Iteration 193, loss = 0.48553505\n",
      "Validation score: 0.740000\n",
      "Iteration 194, loss = 0.48482572\n",
      "Validation score: 0.740000\n",
      "Iteration 195, loss = 0.48446790\n",
      "Validation score: 0.740000\n",
      "Iteration 196, loss = 0.48370578\n",
      "Validation score: 0.740000\n",
      "Iteration 197, loss = 0.48298694\n",
      "Validation score: 0.760000\n",
      "Iteration 198, loss = 0.48236070\n",
      "Validation score: 0.780000\n",
      "Iteration 199, loss = 0.48167016\n",
      "Validation score: 0.780000\n",
      "Iteration 200, loss = 0.48102867\n",
      "Validation score: 0.780000\n",
      "Iteration 201, loss = 0.48046496\n",
      "Validation score: 0.780000\n",
      "Iteration 202, loss = 0.48019366\n",
      "Validation score: 0.780000\n",
      "Iteration 203, loss = 0.47949193\n",
      "Validation score: 0.780000\n",
      "Iteration 204, loss = 0.47890428\n",
      "Validation score: 0.780000\n",
      "Iteration 205, loss = 0.47825028\n",
      "Validation score: 0.780000\n",
      "Iteration 206, loss = 0.47752812\n",
      "Validation score: 0.780000\n",
      "Iteration 207, loss = 0.47716027\n",
      "Validation score: 0.780000\n",
      "Iteration 208, loss = 0.47652233\n",
      "Validation score: 0.780000\n",
      "Iteration 209, loss = 0.47588591\n",
      "Validation score: 0.780000\n",
      "Iteration 210, loss = 0.47536249\n",
      "Validation score: 0.780000\n",
      "Iteration 211, loss = 0.47490564\n",
      "Validation score: 0.780000\n",
      "Iteration 212, loss = 0.47441336\n",
      "Validation score: 0.780000\n",
      "Iteration 213, loss = 0.47388017\n",
      "Validation score: 0.780000\n",
      "Iteration 214, loss = 0.47337546\n",
      "Validation score: 0.780000\n",
      "Iteration 215, loss = 0.47281153\n",
      "Validation score: 0.780000\n",
      "Iteration 216, loss = 0.47225282\n",
      "Validation score: 0.780000\n",
      "Iteration 217, loss = 0.47150852\n",
      "Validation score: 0.780000\n",
      "Iteration 218, loss = 0.47096839\n",
      "Validation score: 0.780000\n",
      "Iteration 219, loss = 0.47084912\n",
      "Validation score: 0.780000\n",
      "Iteration 220, loss = 0.47037546\n",
      "Validation score: 0.780000\n",
      "Iteration 221, loss = 0.46991456\n",
      "Validation score: 0.780000\n",
      "Iteration 222, loss = 0.46947789\n",
      "Validation score: 0.780000\n",
      "Iteration 223, loss = 0.46899343\n",
      "Validation score: 0.780000\n",
      "Iteration 224, loss = 0.46842632\n",
      "Validation score: 0.780000\n",
      "Iteration 225, loss = 0.46793547\n",
      "Validation score: 0.780000\n",
      "Iteration 226, loss = 0.46750451\n",
      "Validation score: 0.780000\n",
      "Iteration 227, loss = 0.46670225\n",
      "Validation score: 0.780000\n",
      "Iteration 228, loss = 0.46605856\n",
      "Validation score: 0.780000\n",
      "Iteration 229, loss = 0.46549844\n",
      "Validation score: 0.780000\n",
      "Iteration 230, loss = 0.46504173\n",
      "Validation score: 0.780000\n",
      "Iteration 231, loss = 0.46488045\n",
      "Validation score: 0.780000\n",
      "Iteration 232, loss = 0.46447098\n",
      "Validation score: 0.780000\n",
      "Iteration 233, loss = 0.46442434\n",
      "Validation score: 0.780000\n",
      "Iteration 234, loss = 0.46391909\n",
      "Validation score: 0.780000\n",
      "Iteration 235, loss = 0.46350366\n",
      "Validation score: 0.780000\n",
      "Iteration 236, loss = 0.46289299\n",
      "Validation score: 0.780000\n",
      "Iteration 237, loss = 0.46240316\n",
      "Validation score: 0.780000\n",
      "Iteration 238, loss = 0.46191928\n",
      "Validation score: 0.780000\n",
      "Iteration 239, loss = 0.46156755\n",
      "Validation score: 0.780000\n",
      "Iteration 240, loss = 0.46081777\n",
      "Validation score: 0.780000\n",
      "Iteration 241, loss = 0.46047759\n",
      "Validation score: 0.780000\n",
      "Iteration 242, loss = 0.45994648\n",
      "Validation score: 0.780000\n",
      "Iteration 243, loss = 0.45941799\n",
      "Validation score: 0.780000\n",
      "Iteration 244, loss = 0.45927339\n",
      "Validation score: 0.780000\n",
      "Iteration 245, loss = 0.45888476\n",
      "Validation score: 0.780000\n",
      "Iteration 246, loss = 0.45857064\n",
      "Validation score: 0.780000\n",
      "Iteration 247, loss = 0.45810104\n",
      "Validation score: 0.780000\n",
      "Iteration 248, loss = 0.45767199\n",
      "Validation score: 0.780000\n",
      "Iteration 249, loss = 0.45719252\n",
      "Validation score: 0.780000\n",
      "Iteration 250, loss = 0.45674002\n",
      "Validation score: 0.780000\n",
      "Iteration 251, loss = 0.45625050\n",
      "Validation score: 0.780000\n",
      "Iteration 252, loss = 0.45588470\n",
      "Validation score: 0.780000\n",
      "Iteration 253, loss = 0.45542658\n",
      "Validation score: 0.780000\n",
      "Iteration 254, loss = 0.45518369\n",
      "Validation score: 0.780000\n",
      "Iteration 255, loss = 0.45494349\n",
      "Validation score: 0.780000\n",
      "Iteration 256, loss = 0.45476333\n",
      "Validation score: 0.780000\n",
      "Iteration 257, loss = 0.45450763\n",
      "Validation score: 0.780000\n",
      "Iteration 258, loss = 0.45399433\n",
      "Validation score: 0.780000\n",
      "Iteration 259, loss = 0.45315520\n",
      "Validation score: 0.780000\n",
      "Iteration 260, loss = 0.45253038\n",
      "Validation score: 0.780000\n",
      "Iteration 261, loss = 0.45215672\n",
      "Validation score: 0.780000\n",
      "Iteration 262, loss = 0.45169253\n",
      "Validation score: 0.780000\n",
      "Iteration 263, loss = 0.45126316\n",
      "Validation score: 0.780000\n",
      "Iteration 264, loss = 0.45095258\n",
      "Validation score: 0.780000\n",
      "Iteration 265, loss = 0.45056517\n",
      "Validation score: 0.780000\n",
      "Iteration 266, loss = 0.45017548\n",
      "Validation score: 0.780000\n",
      "Iteration 267, loss = 0.44992221\n",
      "Validation score: 0.780000\n",
      "Iteration 268, loss = 0.44950724\n",
      "Validation score: 0.780000\n",
      "Iteration 269, loss = 0.44939081\n",
      "Validation score: 0.780000\n",
      "Iteration 270, loss = 0.44912306\n",
      "Validation score: 0.780000\n",
      "Iteration 271, loss = 0.44862385\n",
      "Validation score: 0.780000\n",
      "Iteration 272, loss = 0.44819795\n",
      "Validation score: 0.780000\n",
      "Iteration 273, loss = 0.44770133\n",
      "Validation score: 0.780000\n",
      "Iteration 274, loss = 0.44735046\n",
      "Validation score: 0.780000\n",
      "Iteration 275, loss = 0.44700839\n",
      "Validation score: 0.780000\n",
      "Iteration 276, loss = 0.44665563\n",
      "Validation score: 0.780000\n",
      "Iteration 277, loss = 0.44636680\n",
      "Validation score: 0.780000\n",
      "Iteration 278, loss = 0.44599474\n",
      "Validation score: 0.780000\n",
      "Iteration 279, loss = 0.44564926\n",
      "Validation score: 0.780000\n",
      "Iteration 280, loss = 0.44528047\n",
      "Validation score: 0.780000\n",
      "Iteration 281, loss = 0.44494875\n",
      "Validation score: 0.780000\n",
      "Iteration 282, loss = 0.44465062\n",
      "Validation score: 0.780000\n",
      "Iteration 283, loss = 0.44435975\n",
      "Validation score: 0.780000\n",
      "Iteration 284, loss = 0.44402768\n",
      "Validation score: 0.780000\n",
      "Iteration 285, loss = 0.44381673\n",
      "Validation score: 0.780000\n",
      "Iteration 286, loss = 0.44343678\n",
      "Validation score: 0.780000\n",
      "Iteration 287, loss = 0.44309463\n",
      "Validation score: 0.780000\n",
      "Iteration 288, loss = 0.44284246\n",
      "Validation score: 0.780000\n",
      "Iteration 289, loss = 0.44260315\n",
      "Validation score: 0.780000\n",
      "Iteration 290, loss = 0.44231415\n",
      "Validation score: 0.780000\n",
      "Iteration 291, loss = 0.44205578\n",
      "Validation score: 0.780000\n",
      "Iteration 292, loss = 0.44175452\n",
      "Validation score: 0.780000\n",
      "Iteration 293, loss = 0.44152201\n",
      "Validation score: 0.780000\n",
      "Iteration 294, loss = 0.44115156\n",
      "Validation score: 0.780000\n",
      "Iteration 295, loss = 0.44083434\n",
      "Validation score: 0.780000\n",
      "Iteration 296, loss = 0.44057590\n",
      "Validation score: 0.780000\n",
      "Iteration 297, loss = 0.44027345\n",
      "Validation score: 0.780000\n",
      "Iteration 298, loss = 0.44000631\n",
      "Validation score: 0.780000\n",
      "Iteration 299, loss = 0.43977009\n",
      "Validation score: 0.780000\n",
      "Validation score did not improve more than tol=0.000100 for 100 consecutive epochs. Stopping.\n"
     ]
    }
   ],
   "source": [
    "# Fitting model based on the train data\n",
    "model = MLPClassifier(random_state=42,\n",
    "                      max_iter=1000, \n",
    "                      hidden_layer_sizes=(10,),\n",
    "                      n_iter_no_change=100,\n",
    "                      early_stopping=True,\n",
    "                      verbose=True).fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0e72d048-ca37-44fa-a774-c89b452fc2b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "loss_curve = model.loss_curve_\n",
    "plt.plot(loss_curve)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "2871dc5b-e1eb-420f-b72a-0f3ea7306824",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiwAAAGdCAYAAAAxCSikAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA3o0lEQVR4nO3df3RU9YH//9ckIZOUH4EQDAFizOIvJOp3TdZsgnRbF9OyH6EcP0r80QgI24MgB4j6LWykGFTSpV0K56thiyZlaVH4fkqx9mtAxyoIjV0rhZYVVmSDDcLENEEy/EhmSHK/f4S5MEwCmXEymZn7fJwzR+bO+07e93rJffH+dW2GYRgCAACIYHH9XQEAAICrIbAAAICIR2ABAAARj8ACAAAiHoEFAABEPAILAACIeAQWAAAQ8QgsAAAg4iX0dwVCpbOzUydOnNDgwYNls9n6uzoAAKAXDMPQ6dOnNWrUKMXF9dyOEjOB5cSJE8rMzOzvagAAgCAcO3ZMY8aM6fHzmAksgwcPltR1wEOGDOnn2gAAgN5wuVzKzMw07+M9iZnA4u0GGjJkCIEFAIAoc7XhHAy6BQAAEY/AAgAAIh6BBQAARDwCCwAAiHgEFgAAEPEILAAAIOIRWAAAQMQjsAAAgIhHYAEAABEvqMBSWVmp7OxsJSUlKTc3V7t37+6x7MyZM2Wz2fxe48eP9ym3Zs0a3XTTTUpOTlZmZqYWL16stra2YKoHAABiTMCBZcuWLVq0aJHKysq0b98+TZw4UZMnT1Z9fX235deuXSun02m+jh07ptTUVD3wwANmmU2bNmnJkiVavny5Dh06pKqqKm3ZskVLly4N/sgAAEDMsBmGYQSyQ35+vu644w6tW7fO3DZu3DhNmzZNFRUVV93/9ddf13333aejR48qKytLkvTEE0/o0KFD+u1vf2uWe/LJJ/Xhhx9esfXmUi6XSykpKWppaeFZQgAARIne3r8Devihx+PR3r17tWTJEp/tRUVFqq2t7dV3VFVVadKkSWZYkaS77rpLv/jFL/Thhx/qzjvvVF1dnWpqajRjxowev8ftdsvtdpvvXS5XIIcCADGj7q9ntOUPx+Tp6OzvqiDGPTYhW5mpX+uXnx1QYGlqalJHR4fS09N9tqenp6uhoeGq+zudTm3fvl2vvvqqz/YHH3xQf/3rX3XXXXfJMAy1t7fr8ccf9wtGl6qoqFB5eXkg1QeAmPSTdz7Vb/50or+rAQuYcvuo6AgsXpc/AtowjKs+FlqSNmzYoKFDh2ratGk+23fu3KkXXnhBlZWVys/P15EjR7Rw4UJlZGRo2bJl3X7X0qVLVVpaar53uVzKzMwM/GAAIMp9edYjSbrnlnTdmD6on2uDWJY+JKnffnZAgSUtLU3x8fF+rSmNjY1+rS6XMwxD1dXVKikpUWJios9ny5YtU0lJiebMmSNJuvXWW3X27Fl973vfU1lZmeLi/McG2+122e32QKoPADGp7XyHJOl/3zFa387J6OfaAH0joFlCiYmJys3NlcPh8NnucDhUWFh4xX137dqlI0eOaPbs2X6fnTt3zi+UxMfHyzAMBTgmGAAsp629K7DYB8T3c02AvhNwl1BpaalKSkqUl5engoICrV+/XvX19Zo7d66krq6a48ePa+PGjT77VVVVKT8/Xzk5OX7fOWXKFK1evVp/+7d/a3YJLVu2TFOnTlV8PH8BAeBK2s53DbZNJrAghgUcWIqLi9Xc3KwVK1bI6XQqJydHNTU15qwfp9PptyZLS0uLtm7dqrVr13b7nc8884xsNpueeeYZHT9+XCNGjNCUKVP0wgsvBHFIAGAtrZ6uFpYkAgtiWMDrsEQq1mEBYFV5zzvUdMajHYsm6uaR/P5DdOnt/ZtnCQFAlPN2CSUl0MKC2EVgAYAo550lRJcQYhmBBQCi2PmOTrV3dvXsM+gWsYzAAgBRzNu6Ikn2AfxKR+zi6gaAKOYdv2KzSfYEfqUjdnF1A0AU87aw2BPievWIFCBaEVgAIIq5L6xyy/gVxDoCCwBEsVbPhSnNBBbEOAILAEQx73OECCyIdQQWAIhil45hAWIZVzgARDFzlVtaWBDjCCwAEMVazzPoFtZAYAGAKHZxWX5+nSO2cYUDQBRz8xwhWASBBQCiGGNYYBUEFgCIYnQJwSq4wgEgirXSJQSLILAAQBSjSwhWQWABgChmrnSbQGBBbCOwAEAUYwwLrIIrHACimDewJCfSwoLYRmABgChmjmGhSwgxjsACAFHMfPghXUKIcVzhABDF2pjWDIsgsABAFGtlWjMsgsACAFHMzdOaYREEFgCIYkxrhlVwhQNAFGtrp0sI1kBgAYAoZrawMK0ZMY7AAgBRyjCMiw8/TOTXOWIbVzgARClPR6cMo+vPdAkh1gUVWCorK5Wdna2kpCTl5uZq9+7dPZadOXOmbDab32v8+PE+5U6dOqX58+crIyNDSUlJGjdunGpqaoKpHgBYgneVW4kuIcS+gAPLli1btGjRIpWVlWnfvn2aOHGiJk+erPr6+m7Lr127Vk6n03wdO3ZMqampeuCBB8wyHo9H99xzjz777DP98pe/1CeffKKXX35Zo0ePDv7IACDGeac0x9mkAfG2fq4N0LcSAt1h9erVmj17tubMmSNJWrNmjd566y2tW7dOFRUVfuVTUlKUkpJivn/99df15ZdfatasWea26upqnTx5UrW1tRowYIAkKSsrK+CDAQArab1klVubjcCC2BZQC4vH49HevXtVVFTks72oqEi1tbW9+o6qqipNmjTJJ5C88cYbKigo0Pz585Wenq6cnBytXLlSHR0dPX6P2+2Wy+XyeQFW4Wnv1Lqd/6Py33ysV3bXqbPT6O8qoR94u4RYNA5WEFALS1NTkzo6OpSenu6zPT09XQ0NDVfd3+l0avv27Xr11Vd9ttfV1endd9/VI488opqaGn366aeaP3++2tvb9YMf/KDb76qoqFB5eXkg1Qdixu5P/6p/3fHf5vv/K3Oo8q5L7ccaoT/wHCFYSVCDbi9vejQMo1fNkRs2bNDQoUM1bdo0n+2dnZ265pprtH79euXm5urBBx9UWVmZ1q1b1+N3LV26VC0tLebr2LFjwRwKEJVOnvX4vP/y3Pl+qgn6E09qhpUE1MKSlpam+Ph4v9aUxsZGv1aXyxmGoerqapWUlCgxMdHns4yMDA0YMEDx8Rf/lTBu3Dg1NDTI4/H4lZcku90uu90eSPWBmOFd3dR8f77n7lPErlYWjYOFBBTLExMTlZubK4fD4bPd4XCosLDwivvu2rVLR44c0ezZs/0+mzBhgo4cOaLOzou/hA8fPqyMjIxuwwpgde7LAgqBxZrazCc108KC2BfwVV5aWqpXXnlF1dXVOnTokBYvXqz6+nrNnTtXUldXzaOPPuq3X1VVlfLz85WTk+P32eOPP67m5mYtXLhQhw8f1ptvvqmVK1dq/vz5QRwSEPsuDyiXt7jAGtztF57UnEgLC2JfwNOai4uL1dzcrBUrVsjpdConJ0c1NTXmrB+n0+m3JktLS4u2bt2qtWvXdvudmZmZevvtt7V48WLddtttGj16tBYuXKjvf//7QRwSEPtaLw8sHlpYrIjnCMFKAg4skjRv3jzNmzev2882bNjgty0lJUXnzp274ncWFBTo97//fTDVASzn0hVOu94TWKzoYpcQgQWxj45PIAr5dwkRWKyolVlCsBCuciAK+bewMIbFirzBlYXjYAUEFiAKeVtUhn2t61EWdAlZE11CsBICCxCFvINsh32ta9r/5YNwYQ0XV7rlVzliH1c5EIW8LSwpF1pY3HQJWRKzhGAlBBYgCnm7AoYm0yVkZTxLCFZCYAGikPdG5e0SYpaQNZljWFg4DhZAYAGikHfMylDvGBYWjrMkb1BNSuBXOWIfVzkQhbxjVoaas4QYw2JFdAnBSggsQBRqM1tYLgQWuoQsqZVpzbAQAgsQhbyBJSWZWUJW5mbhOFgIgQWIMoZhmE9nNgfdMkvIkliHBVbCVQ5EmfMdhjo6DUksHGd1rHQLKyGwAFHm0vEqQy9Zmt8wjP6qEvpJKy0ssBCuciDKeLsBbDZpSFJXYOk0ulpeYC3ea8HOSrewAAILEGW8A2yTEuKVlHjxrzAzhazFMAy5L4xlSmbhOFgAgQWIMpcOtEyMj5PN5rsd1uANKxJjWGANBBYgyrResliYzWYzH3zX5mFqs5VcuroxK93CCrjKgShz+cwQ74BLuoSsxfv/OyHOpoR4fpUj9nGVA1Hm4kDLrr++3uBCl5C1MKUZVkNgAaKMN5h4B1omm4GFLiEr4TlCsBoCCxBlzDEsF8au2C/csFg8zlpY5RZWw5UORBlzWvOAOJ//0iVkLa20sMBiCCxAlPEOtjQH3SYwhsWKLg+uQKzjSgeiTNtlT+j1jmXhic3Wcvl1AMQ6AgsQZbyDa+1Ma7a0y1vagFhHYAGizOUPvPN2CV26kBhiX+uFhQJ5jhCsgsACRJnLp7PamdZsScwSgtVwpQNRpu2Shx9KdAlZFV1CsBoCCxBl3ObCcV1/fZNZ6daSvMGVQbewCgILEGUuX3+DpfmtiS4hWE1QV3plZaWys7OVlJSk3Nxc7d69u8eyM2fOlM1m83uNHz++2/KbN2+WzWbTtGnTgqkaEPPaLlvp9uLCcYxhsRKW5ofVBBxYtmzZokWLFqmsrEz79u3TxIkTNXnyZNXX13dbfu3atXI6nebr2LFjSk1N1QMPPOBX9i9/+YueeuopTZw4MfAjASzi4rRmHn5oZQQWWE3AgWX16tWaPXu25syZo3HjxmnNmjXKzMzUunXrui2fkpKikSNHmq+PPvpIX375pWbNmuVTrqOjQ4888ojKy8v1N3/zN8EdDWAB3sGWyXQJWRpPa4bVJARS2OPxaO/evVqyZInP9qKiItXW1vbqO6qqqjRp0iRlZWX5bF+xYoVGjBih2bNnX7GLycvtdsvtdpvvXS5Xr34+EMlqjzTJceiLK5apbz4nyX8My383nFb5bz7u2woiYvzp81OSGMMC6wgosDQ1Namjo0Pp6ek+29PT09XQ0HDV/Z1Op7Zv365XX33VZ/vvfvc7VVVVaf/+/b2uS0VFhcrLy3tdHogGT/6fP8nZ0tarsqkDEyVJaRf+62xp089+91lfVQ0RaviF//9ArAsosHjZbDaf94Zh+G3rzoYNGzR06FCfAbWnT5/Wd7/7Xb388stKS0vrdR2WLl2q0tJS873L5VJmZmav9wci0cmzHknSowVZGpzU81/Pa1O/pvGjhkiS8v9muJ6blqOGltaw1BGRI22QXXffnH71gkAMCCiwpKWlKT4+3q81pbGx0a/V5XKGYai6ulolJSVKTLz4L4L/+Z//0WeffaYpU6aY2zo7u/pmExIS9Mknn2js2LF+32e322W32wOpPhDRDMOQu73r2l9w9w0aMbh313d8nE0lf5919YIAEMUC6vxMTExUbm6uHA6Hz3aHw6HCwsIr7rtr1y4dOXJEs2fP9tl+880368CBA9q/f7/5mjp1qr75zW9q//79tJrAMrxhRWJcAgBcLuAuodLSUpWUlCgvL08FBQVav3696uvrNXfuXEldXTXHjx/Xxo0bffarqqpSfn6+cnJyfLYnJSX5bRs6dKgk+W0HYtmls3yY+QEAvgIOLMXFxWpubtaKFSvkdDqVk5Ojmpoac9aP0+n0W5OlpaVFW7du1dq1a0NTayAGeVewTYizaUA8LSwAcCmbYRhGf1ciFFwul1JSUtTS0qIhQ4b0d3WAgB1tOqtv/ninBtkT9F/l3+rv6gBAWPT2/s0/44AIwbNhAKBn/GYEIoQ3sNgTGL8CAJcjsAARwrvUenIigQUALkdgASIEXUIA0DN+MwIRwgwsdAkBgB8CCxAhvE9hZg0WAPBHYAEihHcMC4EFAPwRWIAI0ephDAsA9ITfjECEoEsIAHpGYAEixMUuIf5aAsDl+M0IRAg3s4QAoEcEFiBCeKc1s3AcAPgjsAARovU8Y1gAoCcEFiBCeMew2BP4awkAl+M3IxAh2mhhAYAeEViACNHWfuHhhwQWAPBDYAEiBC0sANAzAgsQIXhaMwD0jN+MQISghQUAekZgASIEK90CQM/4zQhECFpYAKBnBBYgQrBwHAD0jMACRAi32SVEYAGAyxFYgAjQ0WnI03EhsLDSLQD44TcjEAHc7R3mn3n4IQD4I7AAEcA7Q0iSkhIILABwOQILEAG8A24T4+MUF2fr59oAQOQhsAARwDul2c4aLADQLX47AhGANVgA4MoILEAE8I5h4UnNANA9AgsQAXjwIQBcWVC/HSsrK5Wdna2kpCTl5uZq9+7dPZadOXOmbDab32v8+PFmmZdfflkTJ07UsGHDNGzYME2aNEkffvhhMFUDohJdQgBwZQEHli1btmjRokUqKyvTvn37NHHiRE2ePFn19fXdll+7dq2cTqf5OnbsmFJTU/XAAw+YZXbu3KmHHnpI7733nj744ANde+21Kioq0vHjx4M/MiCKmA8+ZEozAHTLZhiGEcgO+fn5uuOOO7Ru3Tpz27hx4zRt2jRVVFRcdf/XX39d9913n44ePaqsrKxuy3R0dGjYsGF68cUX9eijj/aqXi6XSykpKWppadGQIUN6dzBAhNi693M9+X/+pK/fOEIbH7uzv6sDAGHT2/t3QiBf6vF4tHfvXi1ZssRne1FRkWpra3v1HVVVVZo0aVKPYUWSzp07p/Pnzys1NbXHMm63W26323zvcrl69fMRPVxt51W956haWs/3d1X63OEvTktiWX4A6ElAgaWpqUkdHR1KT0/32Z6enq6Ghoar7u90OrV9+3a9+uqrVyy3ZMkSjR49WpMmTeqxTEVFhcrLy3tXcUSlX+8/oTXvfNrf1Qir4YMS+7sKABCRAgosXjab70qchmH4bevOhg0bNHToUE2bNq3HMqtWrdJrr72mnTt3KikpqcdyS5cuVWlpqfne5XIpMzPz6pVH1PjyrEeSNH7UEH3jphH9XJu+Z0+I1wN5Y/q7GgAQkQIKLGlpaYqPj/drTWlsbPRrdbmcYRiqrq5WSUmJEhO7/1fkj3/8Y61cuVLvvPOObrvttit+n91ul91uD6T6iDLemTN3Zqfq6W/d3M+1AQD0p4A6zBMTE5WbmyuHw+Gz3eFwqLCw8Ir77tq1S0eOHNHs2bO7/fxHP/qRnnvuOe3YsUN5eXmBVAsxypw5w1RfALC8gLuESktLVVJSory8PBUUFGj9+vWqr6/X3LlzJXV11Rw/flwbN2702a+qqkr5+fnKycnx+85Vq1Zp2bJlevXVV3XdddeZLTiDBg3SoEGDgjkuxIC29q4WFlZ/BQAEHFiKi4vV3NysFStWyOl0KicnRzU1NeasH6fT6bcmS0tLi7Zu3aq1a9d2+52VlZXyeDy6//77fbYvX75czz77bKBVRIxo87D6KwCgS1CDbufNm6d58+Z1+9mGDRv8tqWkpOjcuXM9ft9nn30WTDUQ47wtLHQJAQD4pysiFqu/AgC8CCyIWN5ZQna6hADA8rgTIGK1nmfQLQCgC4EFEYtpzQAALwILIpb7PINuAQBdCCyIWG3nmdYMAOjCnQARizEsAAAvAgsiFmNYAABeBBZEJMMwzIXjmNYMAOBOgIjk6eiUYXT9mRYWAACBBRGpzdNp/pmVbgEABBZEJG93UHycTQPibf1cGwBAfyOwICKZU5oT4mSzEVgAwOoILIhIzBACAFyKwIKI1MoqtwCASxBYEJFY5RYAcCnuBohIbbSwAAAuQWBBRGIMCwDgUgQWRCS6hAAAl+JugIh0cVozLSwAAAILIpQZWBIJLAAAAgsiVFv7hTEstLAAAERgQYRq9TCGBQBwEXcDRCTvs4SYJQQAkAgsiFDuC9OakwksAAARWBChmNYMALgUdwNEJFa6BQBcisCCiOR9+KGdwAIAEIEFEcpcmj+BSxQAQGBBhPJ2CSWzcBwAQAQWRCgWjgMAXCqowFJZWans7GwlJSUpNzdXu3fv7rHszJkzZbPZ/F7jx4/3Kbd161bdcsststvtuuWWW7Rt27ZgqoYY0eZh0C0A4KKAA8uWLVu0aNEilZWVad++fZo4caImT56s+vr6bsuvXbtWTqfTfB07dkypqal64IEHzDIffPCBiouLVVJSoj/96U8qKSnR9OnT9Z//+Z/BHxmi2sWF42gEBABINsMwjEB2yM/P1x133KF169aZ28aNG6dp06apoqLiqvu//vrruu+++3T06FFlZWVJkoqLi+VyubR9+3az3Le//W0NGzZMr732Wq/q5XK5lJKSopaWFg0ZMiSQQ0IEyl/5jr5wufX/LbhLOaNT+rs6AIA+0tv7d0IgX+rxeLR3714tWbLEZ3tRUZFqa2t79R1VVVWaNGmSGVakrhaWxYsX+5T71re+pTVr1vT4PW63W26323zvcrl69fOj2c5PGrXr8F/7uxphcerceUl0CQEAugQUWJqamtTR0aH09HSf7enp6WpoaLjq/k6nU9u3b9err77qs72hoSHg76yoqFB5eXkAtY9+C17bp9Nt7f1djbCx2aShXxvQ39UAAESAgAKLl81m83lvGIbftu5s2LBBQ4cO1bRp077ydy5dulSlpaXme5fLpczMzKvWIVq1d3SaYWXOXdmyW2BsR86oFKUNsvd3NQAAESCgwJKWlqb4+Hi/lo/Gxka/FpLLGYah6upqlZSUKDEx0eezkSNHBvyddrtddrt1bmbeab6S9GTRTaxPAgCwlID+mZ6YmKjc3Fw5HA6f7Q6HQ4WFhVfcd9euXTpy5Ihmz57t91lBQYHfd7799ttX/U4r8S6kJkl2Vn8FAFhMwF1CpaWlKikpUV5engoKCrR+/XrV19dr7ty5krq6ao4fP66NGzf67FdVVaX8/Hzl5OT4fefChQv19a9/Xf/6r/+q73znO/r1r3+td955R3v27AnysGKPN7DYE+IUF3f17jcAAGJJwIGluLhYzc3NWrFihZxOp3JyclRTU2PO+nE6nX5rsrS0tGjr1q1au3Ztt99ZWFiozZs365lnntGyZcs0duxYbdmyRfn5+UEcUmzi6cUAACsLeB2WSBXr67D81/EW3fv/7FH6ELv+818m9Xd1AAAIid7evxkMESVoYQEAWBmBJUq0nedhgAAA6yKwRAmzhYXpzAAACyKwRIlWb2BhSjMAwIK4+0UJxrAAAKyMwBIlvCvdJllgSX4AAC7H3S9KuC+0sCTTwgIAsCACS5Ro9dAlBACwLgJLlGhrJ7AAAKyLwBIlvOuw2BnDAgCwIO5+UcKcJcTCcQAACyKwRAlvC0syC8cBACyIwBIl2lg4DgBgYdz9ogQLxwEArIzAEiWYJQQAsDICS5Qwn9ZMYAEAWBCBJUpc7BLifxkAwHq4+0WJVsawAAAsjMASJdx0CQEALIzAEiXoEgIAWBl3vyjRxtOaAQAWRmCJAoZhMIYFAGBpBJYocL7DUKfR9WeeJQQAsCICSxTwLhon8bRmAIA1cfeLAt7xKzabZOdZQgAAC+LuFwXMKc0J8bLZbP1cGwAAwo/AEgVamdIMALA47oBRgCc1AwCsjsASBXjwIQDA6ggsUYAWFgCA1RFYogBjWAAAVhfUHbCyslLZ2dlKSkpSbm6udu/efcXybrdbZWVlysrKkt1u19ixY1VdXe1TZs2aNbrpppuUnJyszMxMLV68WG1tbcFUL+aYLSwsGgcAsKiEQHfYsmWLFi1apMrKSk2YMEE//elPNXnyZB08eFDXXnttt/tMnz5dX3zxhaqqqnT99dersbFR7e3t5uebNm3SkiVLVF1drcLCQh0+fFgzZ86UJP3kJz8J7shiyMUnNdPCAgCwpoADy+rVqzV79mzNmTNHUlfLyFtvvaV169apoqLCr/yOHTu0a9cu1dXVKTU1VZJ03XXX+ZT54IMPNGHCBD388MPm5w899JA+/PDDQKsXk7wr3SYn0sICALCmgAKLx+PR3r17tWTJEp/tRUVFqq2t7XafN954Q3l5eVq1apV+/vOfa+DAgZo6daqee+45JScnS5Luuusu/eIXv9CHH36oO++8U3V1daqpqdGMGTN6rIvb7Zbb7Tbfu1yuQA6lz235Q73+u+F0SL7r4+Ndx0aXEADAqgIKLE1NTero6FB6errP9vT0dDU0NHS7T11dnfbs2aOkpCRt27ZNTU1Nmjdvnk6ePGmOY3nwwQf117/+VXfddZcMw1B7e7sef/xxv2B0qYqKCpWXlwdS/bA5dvKcvr/1QMi/N3VgYsi/EwCAaBBwl5Akv+XhDcPoccn4zs5O2Ww2bdq0SSkpKZK6upXuv/9+vfTSS0pOTtbOnTv1wgsvqLKyUvn5+Tpy5IgWLlyojIwMLVu2rNvvXbp0qUpLS833LpdLmZmZwRxOyJ06d16SNMieoBmFWSH5zuQB8Zr+d5FxfAAAhFtAgSUtLU3x8fF+rSmNjY1+rS5eGRkZGj16tBlWJGncuHEyDEOff/65brjhBi1btkwlJSXmuJhbb71VZ8+e1fe+9z2VlZUpLs5/sKndbpfdbg+k+mHjHXOSNihRT3/r5n6uDQAA0S+gaSeJiYnKzc2Vw+Hw2e5wOFRYWNjtPhMmTNCJEyd05swZc9vhw4cVFxenMWPGSJLOnTvnF0ri4+NlGIYMwwikihGBhd4AAAitgOfJlpaW6pVXXlF1dbUOHTqkxYsXq76+XnPnzpXU1VXz6KOPmuUffvhhDR8+XLNmzdLBgwf1/vvv6+mnn9Zjjz1mDrqdMmWK1q1bp82bN+vo0aNyOBxatmyZpk6dqvj46Lvps5Q+AAChFfAYluLiYjU3N2vFihVyOp3KyclRTU2NsrK6xmo4nU7V19eb5QcNGiSHw6EFCxYoLy9Pw4cP1/Tp0/X888+bZZ555hnZbDY988wzOn78uEaMGKEpU6bohRdeCMEhhh8r0wIAEFo2Ixr7XLrhcrmUkpKilpYWDRkypF/r8v9+dEz/9y//rG/cNEIbZt3Zr3UBACCS9fb+TRNAH3CzlD4AACFFYOkDbSylDwBASHFH7QPeWUIspQ8AQGgQWPqAd9CtnS4hAABCgsDSB5jWDABAaBFY+oB3pVvGsAAAEBrcUfuAOYaFFhYAAEKCwNIHWJofAIDQIrD0AaY1AwAQWtxR+wAtLAAAhBaBpQ+0Ma0ZAICQIrD0AW+XEAvHAQAQGgSWPmB2CSVwegEACAXuqH2AMSwAAIQWgaUPtLWz0i0AAKFEYOkDLBwHAEBoEVhCzDCMS7qEOL0AAIQCd9QQ83R0qtPo+rOdFhYAAEKCwBJi3inNEi0sAACECnfUEHNf6A6y2aTEeE4vAAChwB01xMxF4wbEy2az9XNtAACIDQSWEGtlDRYAAEKOwBJirHILAEDocVcNMVa5BQAg9AgsIcYqtwAAhB6BJcRYNA4AgNDjrhpidAkBABB6BJYQI7AAABB6BJYQ867DQpcQAAChw101xGhhAQAg9AgsIcbCcQAAhF5QgaWyslLZ2dlKSkpSbm6udu/efcXybrdbZWVlysrKkt1u19ixY1VdXe1T5tSpU5o/f74yMjKUlJSkcePGqaamJpjq9SuzSyiBwAIAQKgkBLrDli1btGjRIlVWVmrChAn66U9/qsmTJ+vgwYO69tpru91n+vTp+uKLL1RVVaXrr79ejY2Nam9vNz/3eDy65557dM011+iXv/ylxowZo2PHjmnw4MHBH1k/YVozAAChF3BgWb16tWbPnq05c+ZIktasWaO33npL69atU0VFhV/5HTt2aNeuXaqrq1Nqaqok6brrrvMpU11drZMnT6q2tlYDBgyQJGVlZQVatYjgbu8KLMl0CQEAEDIBNQN4PB7t3btXRUVFPtuLiopUW1vb7T5vvPGG8vLytGrVKo0ePVo33nijnnrqKbW2tvqUKSgo0Pz585Wenq6cnBytXLlSHR0dPdbF7XbL5XL5vMLNMAz9/IPPVP6bj83XHz77UhJjWAAACKWAWliamprU0dGh9PR0n+3p6elqaGjodp+6ujrt2bNHSUlJ2rZtm5qamjRv3jydPHnSHMdSV1end999V4888ohqamr06aefav78+Wpvb9cPfvCDbr+3oqJC5eXlgVQ/5A5/cUbLfv1xt5+lDkwMc20AAIhdAXcJSZLNZvN5bxiG3zavzs5O2Ww2bdq0SSkpKZK6upXuv/9+vfTSS0pOTlZnZ6euueYarV+/XvHx8crNzdWJEyf0ox/9qMfAsnTpUpWWlprvXS6XMjMzgzmcoJ0865EkDfvaAD2cf3H8zrCvJep/3ZYR1roAABDLAgosaWlpio+P92tNaWxs9Gt18crIyNDo0aPNsCJJ48aNk2EY+vzzz3XDDTcoIyNDAwYMUHx8vE+ZhoYGeTweJSb6t1bY7XbZ7fZAqh9ybRfGq4wamqynv3Vzv9YFAIBYFtAYlsTEROXm5srhcPhsdzgcKiws7HafCRMm6MSJEzpz5oy57fDhw4qLi9OYMWPMMkeOHFFnZ6dPmYyMjG7DSqRws+YKAABhEfDc29LSUr3yyiuqrq7WoUOHtHjxYtXX12vu3LmSurpqHn30UbP8ww8/rOHDh2vWrFk6ePCg3n//fT399NN67LHHlJycLEl6/PHH1dzcrIULF+rw4cN68803tXLlSs2fPz9Eh9k3vGuuMCMIAIC+FfAYluLiYjU3N2vFihVyOp3KyclRTU2NOQ3Z6XSqvr7eLD9o0CA5HA4tWLBAeXl5Gj58uKZPn67nn3/eLJOZmam3335bixcv1m233abRo0dr4cKF+v73vx+CQ+w7ray5AgBAWNgMwzD6uxKh4HK5lJKSopaWFg0ZMiQsP/Nnvzuq8t8c1P+6LUMvPXxHWH4mAACxpLf3b5oGvgKW4QcAIDwILF+Bdxn+5EROIwAAfYk77VfgndZMCwsAAH2LwPIVtHmY1gwAQDgQWL4CcwwLs4QAAOhT3Gm/ArNLiBYWAAD6FIHlK2hjpVsAAMKCwPIVtJpdQgQWAAD6EoHlK2hjpVsAAMKCO+1XYD78kGnNAAD0KQLLV2A+/DCRwAIAQF8isHwFPPwQAIDw4E77FXjHsNjpEgIAoE8RWL4CpjUDABAeBJavoK2dMSwAAIQDgSVInZ2GPBcCS1ICpxEAgL7EnTZI3mX5JbqEAADoawSWIHmnNEsEFgAA+hqBJUjeAbcD4m2Kj7P1c20AAIhtBJYgMUMIAIDwIbAEqZXAAgBA2BBYgtRmPqmZUwgAQF/jbhskHnwIAED4EFiC5J3WzKJxAAD0PQJLkFo93kXjCCwAAPQ1AkuQzAcfMoYFAIA+x902SN4uIWYJAQDQ9wgsQfLOEkomsAAA0OcILEG6uHAcpxAAgL7G3TZIrHQLAED4EFiCRGABACB8ggoslZWVys7OVlJSknJzc7V79+4rlne73SorK1NWVpbsdrvGjh2r6urqbstu3rxZNptN06ZNC6ZqYWOudJtA5gMAoK8lBLrDli1btGjRIlVWVmrChAn66U9/qsmTJ+vgwYO69tpru91n+vTp+uKLL1RVVaXrr79ejY2Nam9v9yv3l7/8RU899ZQmTpwY+JGEmdnCwsJxAAD0uYADy+rVqzV79mzNmTNHkrRmzRq99dZbWrdunSoqKvzK79ixQ7t27VJdXZ1SU1MlSdddd51fuY6ODj3yyCMqLy/X7t27derUqUCrFlatLM0PAEDYBNSf4fF4tHfvXhUVFflsLyoqUm1tbbf7vPHGG8rLy9OqVas0evRo3XjjjXrqqafU2trqU27FihUaMWKEZs+e3au6uN1uuVwun1c4XXz4IYEFAIC+FlALS1NTkzo6OpSenu6zPT09XQ0NDd3uU1dXpz179igpKUnbtm1TU1OT5s2bp5MnT5rjWH73u9+pqqpK+/fv73VdKioqVF5eHkj1Q8rdzrRmAADCJai7rc1m83lvGIbfNq/Ozk7ZbDZt2rRJd955p/7pn/5Jq1ev1oYNG9Ta2qrTp0/ru9/9rl5++WWlpaX1ug5Lly5VS0uL+Tp27FgwhxI07xgWFo4DAKDvBdTCkpaWpvj4eL/WlMbGRr9WF6+MjAyNHj1aKSkp5rZx48bJMAx9/vnnOnv2rD777DNNmTLF/Lyzs6u7JSEhQZ988onGjh3r9712u112uz2Q6odUK9OaAQAIm4BaWBITE5WbmyuHw+Gz3eFwqLCwsNt9JkyYoBMnTujMmTPmtsOHDysuLk5jxozRzTffrAMHDmj//v3ma+rUqfrmN7+p/fv3KzMzM4jD6nveMSw8/BAAgL4X8Cyh0tJSlZSUKC8vTwUFBVq/fr3q6+s1d+5cSV1dNcePH9fGjRslSQ8//LCee+45zZo1S+Xl5WpqatLTTz+txx57TMnJyZKknJwcn58xdOjQbrdHEhaOAwAgfAIOLMXFxWpubtaKFSvkdDqVk5OjmpoaZWVlSZKcTqfq6+vN8oMGDZLD4dCCBQuUl5en4cOHa/r06Xr++edDdxT94OLCcQQWAAD6ms0wDKO/KxEKLpdLKSkpamlp0ZAhQ/r85926/C2ddrfrvae+oey0gX3+8wAAiEW9vX8zACNIrTytGQCAsOFuG4T2jk61d3Y1TNElBABA3yOwBKGtvdP8M4NuAQDoewSWIHhnCEmSnac1AwDQ57jbBqHV0xVY7AlxiovrfoVfAAAQOgSWIFx8jhDdQQAAhAOBJQgXn9TM6QMAIBy44waBBx8CABBeBJYgXGxhIbAAABAOBJYgeBeNsxNYAAAICwJLEMwHHzKlGQCAsOCOGwSe1AwAQHgRWILgXemWQbcAAIQHgSUIbR4efAgAQDhxxw0CXUIAAIQXgSUIbax0CwBAWBFYgsA6LAAAhBeBJQit5xnDAgBAOHHHDQJjWAAACC8CSxDc3i4hFo4DACAsuOMGwXz4YSItLAAAhAOBJQjMEgIAILwILEFovbBwnD2BwAIAQDgQWIJwcVozpw8AgHDgjhsEuoQAAAgvAksQvLOEePghAADhQWAJQivrsAAAEFYEliC0sdItAABhxR03QIZhsNItAABhRmAJ0PkOQ51G158JLAAAhAeBJUDe8SsSXUIAAIRLUHfcyspKZWdnKykpSbm5udq9e/cVy7vdbpWVlSkrK0t2u11jx45VdXW1+fnLL7+siRMnatiwYRo2bJgmTZqkDz/8MJiq9Tn3hcBis0mJ8QQWAADCIeA77pYtW7Ro0SKVlZVp3759mjhxoiZPnqz6+voe95k+fbp++9vfqqqqSp988olee+013XzzzebnO3fu1EMPPaT33ntPH3zwga699loVFRXp+PHjwR1VHzIXjUuIl81m6+faAABgDTbDMIxAdsjPz9cdd9yhdevWmdvGjRunadOmqaKiwq/8jh079OCDD6qurk6pqam9+hkdHR0aNmyYXnzxRT366KO92sflciklJUUtLS0aMmRI7w4mCIe/OK2in7yv1IGJ+uOye/rs5wAAYAW9vX8nBPKlHo9He/fu1ZIlS3y2FxUVqba2ttt93njjDeXl5WnVqlX6+c9/roEDB2rq1Kl67rnnlJyc3O0+586d0/nz568YcNxut9xut/ne5XIFcii9VrXnqD7/8pz5vvmMR5KUlEB3EAAA4RJQYGlqalJHR4fS09N9tqenp6uhoaHbferq6rRnzx4lJSVp27Ztampq0rx583Ty5EmfcSyXWrJkiUaPHq1Jkyb1WJeKigqVl5cHUv2gvPnnE/pj/Sm/7amDEvv8ZwMAgC4BBRavy8duGIbR43iOzs5O2Ww2bdq0SSkpKZKk1atX6/7779dLL73k18qyatUqvfbaa9q5c6eSkpJ6rMPSpUtVWlpqvne5XMrMzAzmcK7of+eOUcHY4T7bbLLp2zkjQ/6zAABA9wIKLGlpaYqPj/drTWlsbPRrdfHKyMjQ6NGjzbAidY15MQxDn3/+uW644QZz+49//GOtXLlS77zzjm677bYr1sVut8tutwdS/aA8kp/V5z8DAABcWUADMRITE5WbmyuHw+Gz3eFwqLCwsNt9JkyYoBMnTujMmTPmtsOHDysuLk5jxowxt/3oRz/Sc889px07digvLy+QagEAgBgX8MjR0tJSvfLKK6qurtahQ4e0ePFi1dfXa+7cuZK6umoundnz8MMPa/jw4Zo1a5YOHjyo999/X08//bQee+wxszto1apVeuaZZ1RdXa3rrrtODQ0Namho8Ak5AADAugIew1JcXKzm5matWLFCTqdTOTk5qqmpUVZWV9eJ0+n0WZNl0KBBcjgcWrBggfLy8jR8+HBNnz5dzz//vFmmsrJSHo9H999/v8/PWr58uZ599tkgDw0AAMSKgNdhiVThWocFAACETm/v3ywmAgAAIh6BBQAARDwCCwAAiHgEFgAAEPEILAAAIOIRWAAAQMQjsAAAgIhHYAEAABGPwAIAACJewEvzRyrvgr0ul6ufawIAAHrLe9++2sL7MRNYTp8+LUnKzMzs55oAAIBAnT59WikpKT1+HjPPEurs7NSJEyc0ePBg2Wy2kH2vy+VSZmamjh07xjOKeoHzFRjOV+9xrgLD+QoM56v3Qn2uDMPQ6dOnNWrUKMXF9TxSJWZaWOLi4jRmzJg++/4hQ4ZwEQeA8xUYzlfvca4Cw/kKDOer90J5rq7UsuLFoFsAABDxCCwAACDiEViuwm63a/ny5bLb7f1dlajA+QoM56v3OFeB4XwFhvPVe/11rmJm0C0AAIhdtLAAAICIR2ABAAARj8ACAAAiHoEFAABEPALLVVRWVio7O1tJSUnKzc3V7t27+7tK/e7ZZ5+VzWbzeY0cOdL83DAMPfvssxo1apSSk5P1jW98Qx9//HE/1ji83n//fU2ZMkWjRo2SzWbT66+/7vN5b86P2+3WggULlJaWpoEDB2rq1Kn6/PPPw3gU4XG1czVz5ky/a+3v//7vfcpY5VxJUkVFhf7u7/5OgwcP1jXXXKNp06bpk08+8SnD9dWlN+eK6+uidevW6bbbbjMXgysoKND27dvNzyPhuiKwXMGWLVu0aNEilZWVad++fZo4caImT56s+vr6/q5avxs/frycTqf5OnDggPnZqlWrtHr1ar344ov6wx/+oJEjR+qee+4xn/cU686ePavbb79dL774Yref9+b8LFq0SNu2bdPmzZu1Z88enTlzRvfee686OjrCdRhhcbVzJUnf/va3fa61mpoan8+tcq4kadeuXZo/f75+//vfy+FwqL29XUVFRTp79qxZhuurS2/OlcT15TVmzBj98Ic/1EcffaSPPvpId999t77zne+YoSQirisDPbrzzjuNuXPn+my7+eabjSVLlvRTjSLD8uXLjdtvv73bzzo7O42RI0caP/zhD81tbW1tRkpKivHv//7vYaph5JBkbNu2zXzfm/Nz6tQpY8CAAcbmzZvNMsePHzfi4uKMHTt2hK3u4Xb5uTIMw5gxY4bxne98p8d9rHquvBobGw1Jxq5duwzD4Pq6ksvPlWFwfV3NsGHDjFdeeSViritaWHrg8Xi0d+9eFRUV+WwvKipSbW1tP9Uqcnz66acaNWqUsrOz9eCDD6qurk6SdPToUTU0NPicN7vdrn/4h3/gvKl352fv3r06f/68T5lRo0YpJyfHkudw586duuaaa3TjjTfqn//5n9XY2Gh+ZvVz1dLSIklKTU2VxPV1JZefKy+uL38dHR3avHmzzp49q4KCgoi5rggsPWhqalJHR4fS09N9tqenp6uhoaGfahUZ8vPztXHjRr311lt6+eWX1dDQoMLCQjU3N5vnhvPWvd6cn4aGBiUmJmrYsGE9lrGKyZMna9OmTXr33Xf1b//2b/rDH/6gu+++W263W5K1z5VhGCotLdVdd92lnJwcSVxfPenuXElcX5c7cOCABg0aJLvdrrlz52rbtm265ZZbIua6ipmnNfcVm83m894wDL9tVjN58mTzz7feeqsKCgo0duxY/cd//Ic5YI3zdmXBnB8rnsPi4mLzzzk5OcrLy1NWVpbefPNN3XfffT3uZ4Vz9cQTT+jPf/6z9uzZ4/cZ15evns4V15evm266Sfv379epU6e0detWzZgxQ7t27TI/7+/rihaWHqSlpSk+Pt4vGTY2NvqlTKsbOHCgbr31Vn366afmbCHOW/d6c35Gjhwpj8ejL7/8sscyVpWRkaGsrCx9+umnkqx7rhYsWKA33nhD7733nsaMGWNu5/ry19O56o7Vr6/ExERdf/31ysvLU0VFhW6//XatXbs2Yq4rAksPEhMTlZubK4fD4bPd4XCosLCwn2oVmdxutw4dOqSMjAxlZ2dr5MiRPufN4/Fo165dnDepV+cnNzdXAwYM8CnjdDr1X//1X5Y/h83NzTp27JgyMjIkWe9cGYahJ554Qr/61a/07rvvKjs72+dzrq+LrnauumP16+tyhmHI7XZHznUVkqG7MWrz5s3GgAEDjKqqKuPgwYPGokWLjIEDBxqfffZZf1etXz355JPGzp07jbq6OuP3v/+9ce+99xqDBw82z8sPf/hDIyUlxfjVr35lHDhwwHjooYeMjIwMw+Vy9XPNw+P06dPGvn37jH379hmSjNWrVxv79u0z/vKXvxiG0bvzM3fuXGPMmDHGO++8Y/zxj3807r77buP222832tvb++uw+sSVztXp06eNJ5980qitrTWOHj1qvPfee0ZBQYExevRoS54rwzCMxx9/3EhJSTF27txpOJ1O83Xu3DmzDNdXl6udK64vX0uXLjXef/994+jRo8af//xn41/+5V+MuLg44+233zYMIzKuKwLLVbz00ktGVlaWkZiYaNxxxx0+U+Ksqri42MjIyDAGDBhgjBo1yrjvvvuMjz/+2Py8s7PTWL58uTFy5EjDbrcbX//6140DBw70Y43D67333jMk+b1mzJhhGEbvzk9ra6vxxBNPGKmpqUZycrJx7733GvX19f1wNH3rSufq3LlzRlFRkTFixAhjwIABxrXXXmvMmDHD7zxY5VwZhtHtuZJk/OxnPzPLcH11udq54vry9dhjj5n3uhEjRhj/+I//aIYVw4iM68pmGIYRmrYaAACAvsEYFgAAEPEILAAAIOIRWAAAQMQjsAAAgIhHYAEAABGPwAIAACIegQUAAEQ8AgsAAIh4BBYAABDxCCwAACDiEVgAAEDEI7AAAICI9/8D91msoeUXtnUAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "validation_scores = model.validation_scores_\n",
    "plt.plot(validation_scores)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "a1afeef0-752c-441c-b813-28ecbf8957b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7983706720977597"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ae671552-d7b3-4630-87fe-49f14cd28277",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7560975609756098"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# checking the score\n",
    "# the result is between 0 and 1: 1 means 100% accurate\n",
    "model.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "78d09e84-9959-42a4-82dd-8dda81a8588a",
   "metadata": {
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([[-0.1599598 ,  0.59600156,  0.17167505, -0.05615316, -0.4140871 ,\n",
       "         -0.46569279,  0.03274084,  0.16719156,  0.14345333,  0.03412946],\n",
       "        [-0.43497265,  0.13785467,  0.27860444, -0.34258153, -0.33909641,\n",
       "         -0.18156969,  0.05340666, -0.07749306,  0.20021612, -0.11901825],\n",
       "        [ 0.05433241, -0.3656574 , -0.19896675, -0.26173755, -0.13551271,\n",
       "          0.22435394,  0.14409416, -0.12040806,  0.11702376, -0.50236871],\n",
       "        [ 0.12693338, -0.25336803, -0.40135614,  0.29650337,  0.46014231,\n",
       "          0.22565946,  0.09462703, -0.3192077 ,  0.15940445, -0.06578116],\n",
       "        [-0.14288605,  0.25134985, -0.39977928,  0.14237605,  0.21212991,\n",
       "          0.04442321, -0.87253147,  0.53632271,  0.55109395,  0.19938205],\n",
       "        [ 0.50948703,  0.42151218,  0.0826315 ,  0.20034846,  0.24185408,\n",
       "          0.40150609, -0.36113525, -0.04642804, -0.23898601, -0.14788759],\n",
       "        [-0.14365134, -0.27124788,  0.27367394, -0.19333193, -0.19858045,\n",
       "         -0.07740582, -0.06232012,  0.23509074, -0.34975447,  0.44502206],\n",
       "        [ 0.26863536,  0.10878642, -0.20968793,  0.24358796,  0.2047712 ,\n",
       "          0.16709734,  0.43364015, -0.40382916, -0.03744276, -0.53512267],\n",
       "        [ 0.34563725, -0.07732675, -0.1696795 , -0.52101497, -0.07813436,\n",
       "         -0.19123032,  0.36541507,  0.19754808,  0.42668626,  0.05131476],\n",
       "        [-0.32447368,  0.28574573,  0.21493489, -0.05626772,  0.39052078,\n",
       "          0.00990733,  0.15394535,  0.03587958, -0.45902504, -0.27511874],\n",
       "        [-0.50184019, -0.07757599, -0.00152181,  0.11942769,  0.13831628,\n",
       "         -0.46504737,  0.31810817,  0.0223644 , -0.30940767, -0.58860403],\n",
       "        [-0.09359612, -0.15091137,  0.3654942 ,  0.08991663,  0.22838776,\n",
       "          0.30689512,  0.1820908 , -0.1399685 ,  0.42637635,  0.15331997],\n",
       "        [ 0.28840369, -0.04505706, -0.00132208, -0.3948559 , -0.18899269,\n",
       "         -0.11388642,  0.3570271 ,  0.33480534,  0.13415735, -0.03664558],\n",
       "        [-0.0802516 , -0.28837918, -0.12819746, -0.20880024,  0.40439554,\n",
       "         -0.15590515,  0.27309984,  0.24796954, -0.07788777,  0.46934092],\n",
       "        [ 0.47196326,  0.00297524, -0.02111532, -0.35453054, -0.03754688,\n",
       "         -0.58891983,  0.14575504, -0.00546613, -0.22215208, -0.31816712],\n",
       "        [ 0.35991288, -0.00532386, -0.34081951, -0.13239223,  0.41424371,\n",
       "         -0.26310266,  0.39679933,  0.18992169, -0.45041492,  0.12702421],\n",
       "        [-0.01020953,  0.06107207,  0.11066135, -0.02204611, -0.14186792,\n",
       "          0.25869869, -0.03884774, -0.04520157,  0.097741  ,  0.31246052],\n",
       "        [ 0.07326449, -0.07598133, -0.00440073, -0.30587835, -0.00129896,\n",
       "         -0.00512821,  0.37417713, -0.26009546,  0.44932007, -0.22942769],\n",
       "        [ 0.0111396 ,  0.04461788,  0.38852725,  0.14367173,  0.03016565,\n",
       "          0.06569028,  0.13740238,  0.34330225,  0.31580551, -0.07897495],\n",
       "        [-0.5080911 ,  0.29532315,  0.33091977,  0.30508773, -0.3291225 ,\n",
       "         -0.30671648,  0.54202338,  0.11737069,  0.26697526,  0.11594686]]),\n",
       " array([[-0.01973471],\n",
       "        [ 0.73410324],\n",
       "        [ 0.20495294],\n",
       "        [-0.2852176 ],\n",
       "        [ 0.31494257],\n",
       "        [-0.50663435],\n",
       "        [-1.00198545],\n",
       "        [ 0.61002981],\n",
       "        [ 0.82196701],\n",
       "        [ 0.56908933]])]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.coefs_"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10f10e83-f852-45d0-b8cc-7c10c5ba80be",
   "metadata": {},
   "source": [
    "## Testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "23bc19c7-69a1-46a1-a143-a383f3430089",
   "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>ApplicantIncome</th>\n",
       "      <th>CoapplicantIncome</th>\n",
       "      <th>LoanAmount</th>\n",
       "      <th>Loan_Amount_Term</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Gender_Female</th>\n",
       "      <th>Gender_Male</th>\n",
       "      <th>Married_No</th>\n",
       "      <th>Married_Yes</th>\n",
       "      <th>Dependents_0</th>\n",
       "      <th>Dependents_1</th>\n",
       "      <th>Dependents_2</th>\n",
       "      <th>Dependents_3+</th>\n",
       "      <th>Education_Graduate</th>\n",
       "      <th>Education_Not Graduate</th>\n",
       "      <th>Self_Employed_No</th>\n",
       "      <th>Self_Employed_Yes</th>\n",
       "      <th>Property_Area_Rural</th>\n",
       "      <th>Property_Area_Semiurban</th>\n",
       "      <th>Property_Area_Urban</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>452</th>\n",
       "      <td>0.046976</td>\n",
       "      <td>0.041592</td>\n",
       "      <td>0.202605</td>\n",
       "      <td>0.743590</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>0.492876</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.855282</td>\n",
       "      <td>0.358974</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>437</th>\n",
       "      <td>0.025516</td>\n",
       "      <td>0.027000</td>\n",
       "      <td>0.172214</td>\n",
       "      <td>0.743590</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     ApplicantIncome  CoapplicantIncome  LoanAmount  Loan_Amount_Term  \\\n",
       "452         0.046976           0.041592    0.202605          0.743590   \n",
       "155         0.492876           0.000000    0.855282          0.358974   \n",
       "437         0.025516           0.027000    0.172214          0.743590   \n",
       "\n",
       "     Credit_History  Gender_Female  Gender_Male  Married_No  Married_Yes  \\\n",
       "452             0.0            0.0          1.0         0.0          1.0   \n",
       "155             0.0            0.0          1.0         0.0          1.0   \n",
       "437             1.0            0.0          1.0         0.0          1.0   \n",
       "\n",
       "     Dependents_0  Dependents_1  Dependents_2  Dependents_3+  \\\n",
       "452           1.0           0.0           0.0            0.0   \n",
       "155           0.0           0.0           0.0            1.0   \n",
       "437           1.0           0.0           0.0            0.0   \n",
       "\n",
       "     Education_Graduate  Education_Not Graduate  Self_Employed_No  \\\n",
       "452                 1.0                     0.0               1.0   \n",
       "155                 1.0                     0.0               1.0   \n",
       "437                 1.0                     0.0               1.0   \n",
       "\n",
       "     Self_Employed_Yes  Property_Area_Rural  Property_Area_Semiurban  \\\n",
       "452                0.0                  1.0                      0.0   \n",
       "155                0.0                  0.0                      1.0   \n",
       "437                0.0                  0.0                      0.0   \n",
       "\n",
       "     Property_Area_Urban  \n",
       "452                  0.0  \n",
       "155                  0.0  \n",
       "437                  1.0  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample_record = X_test.sample(3)\n",
    "sample_record"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "76ef2d42-50a0-44ee-bf8a-56a91f9e990d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 1], dtype=int64)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result = model.predict(sample_record)\n",
    "result"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee5d8dc0-1024-4feb-9844-ce99c3234bbb",
   "metadata": {},
   "source": [
    "## Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "db441376-7b4a-4cbb-b5f5-6b503d78f061",
   "metadata": {},
   "source": [
    "### Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "aa498db3-43d6-4f3f-b61f-03e5ebd6a5b7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[14, 29],\n",
       "       [ 1, 79]], dtype=int64)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions = model.predict(X_test)\n",
    "cm = confusion_matrix(y_test, predictions, labels=model.classes_)\n",
    "cm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "98c28e55-ee39-4d89-8da2-caafff9d8e6a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=model.classes_)\n",
    "disp.plot()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c45e1543-1830-4eed-bca2-4c5d78178f1b",
   "metadata": {},
   "source": [
    "### Error Type I & II"
   ]
  },
  {
   "cell_type": "raw",
   "id": "bbf8e47c-d144-44a6-bf4c-4d7bc6ef0acd",
   "metadata": {},
   "source": [
    "True Positive: Correctly detected positive (Actual: 1, Predicted: 1)\n",
    "True Negative: Correctly detected negative (Actual: 0, Predicted: 0)\n",
    "\n",
    "False Positive (Error Type I): Incorrected detected positive (Actual: 0, Predicted: 1)\n",
    "False Negative (Error Type II): Incorrected detected negative (Actual: 1, Predicted: 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4067ca2-f80d-4249-bcb4-95ad7b27285a",
   "metadata": {},
   "source": [
    "### Precision & Recall"
   ]
  },
  {
   "cell_type": "raw",
   "id": "2c4a0405-631d-4146-b7fd-87705a52c857",
   "metadata": {},
   "source": [
    "Precision = TP / (TP + FP)\n",
    "Recall = TP / (TP + FN)\n",
    "\n",
    "Precision = 79 / (79 + 29) = 79/108 = 0.73\n",
    "Recall = 79 / (79 + 1) = 0.987"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1bb0c2a-a849-4932-addc-2fd81d9f6f91",
   "metadata": {},
   "source": [
    "### F1 Score"
   ]
  },
  {
   "cell_type": "raw",
   "id": "7ee4dfc3-3287-4586-9ffc-5975475549a2",
   "metadata": {},
   "source": [
    "f1 = 2 * ((precision * recall)/(precision + recall))\n",
    "f1 = (2 * tp) / (2*tp + fp + fn)\n",
    "\n",
    "f1 = 2 * (0.73 * 0.98) / (0.73 + 0.98) = 2 * (0.715 / 1.71) = 2 * 0.418 = 0.836"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "db545467-14b1-4cc3-8ea4-24ce41e84a6c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8404255319148937"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f1 = f1_score(y_test, predictions)\n",
    "f1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60378132-a84a-4aba-a8f0-d2984ae694d1",
   "metadata": {},
   "source": [
    "## Saving Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "2d8b2b24-46e6-4808-b651-1d8d3931ca32",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['ApplicantIncome', 'CoapplicantIncome', 'LoanAmount',\n",
       "       'Loan_Amount_Term', 'Credit_History', 'Gender_Female',\n",
       "       'Gender_Male', 'Married_No', 'Married_Yes', 'Dependents_0',\n",
       "       'Dependents_1', 'Dependents_2', 'Dependents_3+',\n",
       "       'Education_Graduate', 'Education_Not Graduate', 'Self_Employed_No',\n",
       "       'Self_Employed_Yes', 'Property_Area_Rural',\n",
       "       'Property_Area_Semiurban', 'Property_Area_Urban'], dtype=object)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.feature_names_in_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "a526bdc9-9a47-4dad-80fd-1cbd07e910c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# saving the encoder\n",
    "pickle.dump(model, open(\"../object/model.pkl\", \"wb\"))"
   ]
  }
 ],
 "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": "3.10.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}