AlphabetSoupCharity_Optimization.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import our dependencies\n",
    "import os\n",
    "import pandas as pd\n",
    "import tensorflow as tf\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
    "from tensorflow.keras.callbacks import ModelCheckpoint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>EIN</th>\n",
       "      <th>NAME</th>\n",
       "      <th>APPLICATION_TYPE</th>\n",
       "      <th>AFFILIATION</th>\n",
       "      <th>CLASSIFICATION</th>\n",
       "      <th>USE_CASE</th>\n",
       "      <th>ORGANIZATION</th>\n",
       "      <th>STATUS</th>\n",
       "      <th>INCOME_AMT</th>\n",
       "      <th>SPECIAL_CONSIDERATIONS</th>\n",
       "      <th>ASK_AMT</th>\n",
       "      <th>IS_SUCCESSFUL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10520599</td>\n",
       "      <td>BLUE KNIGHTS MOTORCYCLE CLUB</td>\n",
       "      <td>T10</td>\n",
       "      <td>Independent</td>\n",
       "      <td>C1000</td>\n",
       "      <td>ProductDev</td>\n",
       "      <td>Association</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>5000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10531628</td>\n",
       "      <td>AMERICAN CHESAPEAKE CLUB CHARITABLE TR</td>\n",
       "      <td>T3</td>\n",
       "      <td>Independent</td>\n",
       "      <td>C2000</td>\n",
       "      <td>Preservation</td>\n",
       "      <td>Co-operative</td>\n",
       "      <td>1</td>\n",
       "      <td>1-9999</td>\n",
       "      <td>N</td>\n",
       "      <td>108590</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10547893</td>\n",
       "      <td>ST CLOUD PROFESSIONAL FIREFIGHTERS</td>\n",
       "      <td>T5</td>\n",
       "      <td>CompanySponsored</td>\n",
       "      <td>C3000</td>\n",
       "      <td>ProductDev</td>\n",
       "      <td>Association</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>5000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10553066</td>\n",
       "      <td>SOUTHSIDE ATHLETIC ASSOCIATION</td>\n",
       "      <td>T3</td>\n",
       "      <td>CompanySponsored</td>\n",
       "      <td>C2000</td>\n",
       "      <td>Preservation</td>\n",
       "      <td>Trust</td>\n",
       "      <td>1</td>\n",
       "      <td>10000-24999</td>\n",
       "      <td>N</td>\n",
       "      <td>6692</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10556103</td>\n",
       "      <td>GENETIC RESEARCH INSTITUTE OF THE DESERT</td>\n",
       "      <td>T3</td>\n",
       "      <td>Independent</td>\n",
       "      <td>C1000</td>\n",
       "      <td>Heathcare</td>\n",
       "      <td>Trust</td>\n",
       "      <td>1</td>\n",
       "      <td>100000-499999</td>\n",
       "      <td>N</td>\n",
       "      <td>142590</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        EIN                                      NAME APPLICATION_TYPE  \\\n",
       "0  10520599              BLUE KNIGHTS MOTORCYCLE CLUB              T10   \n",
       "1  10531628    AMERICAN CHESAPEAKE CLUB CHARITABLE TR               T3   \n",
       "2  10547893        ST CLOUD PROFESSIONAL FIREFIGHTERS               T5   \n",
       "3  10553066            SOUTHSIDE ATHLETIC ASSOCIATION               T3   \n",
       "4  10556103  GENETIC RESEARCH INSTITUTE OF THE DESERT               T3   \n",
       "\n",
       "        AFFILIATION CLASSIFICATION      USE_CASE  ORGANIZATION  STATUS  \\\n",
       "0       Independent          C1000    ProductDev   Association       1   \n",
       "1       Independent          C2000  Preservation  Co-operative       1   \n",
       "2  CompanySponsored          C3000    ProductDev   Association       1   \n",
       "3  CompanySponsored          C2000  Preservation         Trust       1   \n",
       "4       Independent          C1000     Heathcare         Trust       1   \n",
       "\n",
       "      INCOME_AMT SPECIAL_CONSIDERATIONS  ASK_AMT  IS_SUCCESSFUL  \n",
       "0              0                      N     5000              1  \n",
       "1         1-9999                      N   108590              1  \n",
       "2              0                      N     5000              0  \n",
       "3    10000-24999                      N     6692              1  \n",
       "4  100000-499999                      N   142590              1  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Import and read the charity_data.csv.\n",
    "application_df = pd.read_csv(\"Resources/charity_data.csv\")\n",
    "application_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>APPLICATION_TYPE</th>\n",
       "      <th>AFFILIATION</th>\n",
       "      <th>CLASSIFICATION</th>\n",
       "      <th>USE_CASE</th>\n",
       "      <th>ORGANIZATION</th>\n",
       "      <th>STATUS</th>\n",
       "      <th>INCOME_AMT</th>\n",
       "      <th>SPECIAL_CONSIDERATIONS</th>\n",
       "      <th>ASK_AMT</th>\n",
       "      <th>IS_SUCCESSFUL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>T10</td>\n",
       "      <td>Independent</td>\n",
       "      <td>C1000</td>\n",
       "      <td>ProductDev</td>\n",
       "      <td>Association</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>5000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>T3</td>\n",
       "      <td>Independent</td>\n",
       "      <td>C2000</td>\n",
       "      <td>Preservation</td>\n",
       "      <td>Co-operative</td>\n",
       "      <td>1</td>\n",
       "      <td>1-9999</td>\n",
       "      <td>N</td>\n",
       "      <td>108590</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>T5</td>\n",
       "      <td>CompanySponsored</td>\n",
       "      <td>C3000</td>\n",
       "      <td>ProductDev</td>\n",
       "      <td>Association</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>N</td>\n",
       "      <td>5000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>T3</td>\n",
       "      <td>CompanySponsored</td>\n",
       "      <td>C2000</td>\n",
       "      <td>Preservation</td>\n",
       "      <td>Trust</td>\n",
       "      <td>1</td>\n",
       "      <td>10000-24999</td>\n",
       "      <td>N</td>\n",
       "      <td>6692</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>T3</td>\n",
       "      <td>Independent</td>\n",
       "      <td>C1000</td>\n",
       "      <td>Heathcare</td>\n",
       "      <td>Trust</td>\n",
       "      <td>1</td>\n",
       "      <td>100000-499999</td>\n",
       "      <td>N</td>\n",
       "      <td>142590</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  APPLICATION_TYPE       AFFILIATION CLASSIFICATION      USE_CASE  \\\n",
       "0              T10       Independent          C1000    ProductDev   \n",
       "1               T3       Independent          C2000  Preservation   \n",
       "2               T5  CompanySponsored          C3000    ProductDev   \n",
       "3               T3  CompanySponsored          C2000  Preservation   \n",
       "4               T3       Independent          C1000     Heathcare   \n",
       "\n",
       "   ORGANIZATION  STATUS     INCOME_AMT SPECIAL_CONSIDERATIONS  ASK_AMT  \\\n",
       "0   Association       1              0                      N     5000   \n",
       "1  Co-operative       1         1-9999                      N   108590   \n",
       "2   Association       1              0                      N     5000   \n",
       "3         Trust       1    10000-24999                      N     6692   \n",
       "4         Trust       1  100000-499999                      N   142590   \n",
       "\n",
       "   IS_SUCCESSFUL  \n",
       "0              1  \n",
       "1              1  \n",
       "2              0  \n",
       "3              1  \n",
       "4              1  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Drop the non-beneficial ID columns, 'EIN' and 'NAME'.\n",
    "application_df = application_df.drop([\"EIN\", \"NAME\"], axis=1)\n",
    "application_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "APPLICATION_TYPE            17\n",
       "AFFILIATION                  6\n",
       "CLASSIFICATION              71\n",
       "USE_CASE                     5\n",
       "ORGANIZATION                 4\n",
       "STATUS                       2\n",
       "INCOME_AMT                   9\n",
       "SPECIAL_CONSIDERATIONS       2\n",
       "ASK_AMT                   8747\n",
       "IS_SUCCESSFUL                2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Determine the number of unique values in each column.\n",
    "application_df.nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ## Adjusting the input data: \n",
    "- to ensure that there are no variables or outliers that are causing confusion in the model\n",
    "    - by bucketing or binning on \n",
    "        - \"APPLICATION_TYPE\" column\n",
    "        - \"CLASSIFICATION\" column\n",
    "        - \"ASK_AMT\" column"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### - - Binning on: \"APPLICATION_TYPE\" column - -"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "T3     27037\n",
       "T4      1542\n",
       "T6      1216\n",
       "T5      1173\n",
       "T19     1065\n",
       "T8       737\n",
       "T7       725\n",
       "T10      528\n",
       "T9       156\n",
       "T13       66\n",
       "T12       27\n",
       "T2        16\n",
       "T25        3\n",
       "T14        3\n",
       "T29        2\n",
       "T15        2\n",
       "T17        1\n",
       "Name: APPLICATION_TYPE, dtype: int64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Look at APPLICATION_TYPE value counts for binning\n",
    "application_counts = application_df.APPLICATION_TYPE.value_counts()\n",
    "application_counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Density'>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the value counts of APPLICATION_TYPE\n",
    "application_counts.plot.density()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "T3       27037\n",
       "T4        1542\n",
       "T6        1216\n",
       "T5        1173\n",
       "T19       1065\n",
       "T8         737\n",
       "T7         725\n",
       "T10        528\n",
       "Other      276\n",
       "Name: APPLICATION_TYPE, dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Determine which values to replace if counts are less than ...?\n",
    "replace_application = list(application_counts[application_counts < 500].index)\n",
    "\n",
    "# Replace in dataframe\n",
    "for app in replace_application:\n",
    "    application_df.APPLICATION_TYPE = application_df.APPLICATION_TYPE.replace(app,\"Other\")\n",
    "    \n",
    "# Check to make sure binning was successful\n",
    "application_df.APPLICATION_TYPE.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### - - Binning on: \"CLASSIFICATION\" column - -"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "C1000    17326\n",
       "C2000     6074\n",
       "C1200     4837\n",
       "C3000     1918\n",
       "C2100     1883\n",
       "         ...  \n",
       "C4120        1\n",
       "C8210        1\n",
       "C2561        1\n",
       "C4500        1\n",
       "C2150        1\n",
       "Name: CLASSIFICATION, Length: 71, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Look at CLASSIFICATION value counts for binning\n",
    "classification_counts = application_df.CLASSIFICATION.value_counts()\n",
    "classification_counts "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Density'>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the value counts of CLASSIFICATION\n",
    "classification_counts.plot.density()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "C1000    17326\n",
       "C2000     6074\n",
       "C1200     4837\n",
       "Other     2261\n",
       "C3000     1918\n",
       "C2100     1883\n",
       "Name: CLASSIFICATION, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Determine which values to replace if counts are less than ..?\n",
    "replace_class = list(classification_counts[classification_counts < 1000].index)\n",
    "\n",
    "# Replace in dataframe\n",
    "for cls in replace_class:\n",
    "    application_df.CLASSIFICATION = application_df.CLASSIFICATION.replace(cls,\"Other\")\n",
    "    \n",
    "# Check to make sure binning was successful\n",
    "application_df.CLASSIFICATION.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### - - Binning on: \"ASK_AMT\" column - -"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5000        25398\n",
       "10478           3\n",
       "15583           3\n",
       "63981           3\n",
       "6725            3\n",
       "            ...  \n",
       "5371754         1\n",
       "30060           1\n",
       "43091152        1\n",
       "18683           1\n",
       "36500179        1\n",
       "Name: ASK_AMT, Length: 8747, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Look at ASK_AMT value counts for binning\n",
    "ask_amt_counts = application_df.ASK_AMT.value_counts()\n",
    "ask_amt_counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='Density'>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the value counts of ASK_AMT\n",
    "ask_amt_counts.plot.density()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Minimum: 5,000\n",
      "Maximum: 8,597,806,340\n"
     ]
    }
   ],
   "source": [
    "# Check minimum and maximum values on \"ASK_AMT\" column\n",
    "print(f\"Minimum: {application_df['ASK_AMT'].min():,}\")\n",
    "print(f\"Maximum: {application_df['ASK_AMT'].max():,}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5000               25398\n",
       "5001-99999          4368\n",
       "100000-999999       2956\n",
       "1000000-9999999     1165\n",
       "10000000+            412\n",
       "Name: ASK_AMT, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Determine which values to replace if counts are less than ..?\n",
    "replace_ask_amt = list(ask_amt_counts[ask_amt_counts <= 5000].index)\n",
    "\n",
    "# Replace in dataframe\n",
    "for amt in replace_ask_amt:\n",
    "    if amt <= 99999:\n",
    "        application_df.ASK_AMT = application_df.ASK_AMT.replace(amt,str(\"5001-99999\"))\n",
    "    elif amt <= 999999:\n",
    "        application_df.ASK_AMT = application_df.ASK_AMT.replace(amt,str(\"100000-999999\"))\n",
    "    elif amt <= 9999999:\n",
    "        application_df.ASK_AMT = application_df.ASK_AMT.replace(amt,str(\"1000000-9999999\"))\n",
    "    else:\n",
    "        application_df.ASK_AMT = application_df.ASK_AMT.replace(amt,str(\"10000000+\"))\n",
    "\n",
    "application_df.ASK_AMT = application_df.ASK_AMT.astype(str)\n",
    "    \n",
    "# Check to make sure binning was successful\n",
    "application_df.ASK_AMT.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "APPLICATION_TYPE          9\n",
       "AFFILIATION               6\n",
       "CLASSIFICATION            6\n",
       "USE_CASE                  5\n",
       "ORGANIZATION              4\n",
       "INCOME_AMT                9\n",
       "SPECIAL_CONSIDERATIONS    2\n",
       "ASK_AMT                   5\n",
       "dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Generate our categorical variable lists\n",
    "application_cat = application_df.dtypes[application_df.dtypes == \"object\"].index.tolist()\n",
    "\n",
    "# Check the number of unique values in each column\n",
    "application_df[application_cat].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>APPLICATION_TYPE_Other</th>\n",
       "      <th>APPLICATION_TYPE_T10</th>\n",
       "      <th>APPLICATION_TYPE_T19</th>\n",
       "      <th>APPLICATION_TYPE_T3</th>\n",
       "      <th>APPLICATION_TYPE_T4</th>\n",
       "      <th>APPLICATION_TYPE_T5</th>\n",
       "      <th>APPLICATION_TYPE_T6</th>\n",
       "      <th>APPLICATION_TYPE_T7</th>\n",
       "      <th>APPLICATION_TYPE_T8</th>\n",
       "      <th>AFFILIATION_CompanySponsored</th>\n",
       "      <th>...</th>\n",
       "      <th>INCOME_AMT_25000-99999</th>\n",
       "      <th>INCOME_AMT_50M+</th>\n",
       "      <th>INCOME_AMT_5M-10M</th>\n",
       "      <th>SPECIAL_CONSIDERATIONS_N</th>\n",
       "      <th>SPECIAL_CONSIDERATIONS_Y</th>\n",
       "      <th>ASK_AMT_100000-999999</th>\n",
       "      <th>ASK_AMT_1000000-9999999</th>\n",
       "      <th>ASK_AMT_10000000+</th>\n",
       "      <th>ASK_AMT_5000</th>\n",
       "      <th>ASK_AMT_5001-99999</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</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>0.0</td>\n",
       "      <td>0.0</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\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>0.0</td>\n",
       "      <td>0.0</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>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>0.0</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 46 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   APPLICATION_TYPE_Other  APPLICATION_TYPE_T10  APPLICATION_TYPE_T19  \\\n",
       "0                     0.0                   1.0                   0.0   \n",
       "1                     0.0                   0.0                   0.0   \n",
       "2                     0.0                   0.0                   0.0   \n",
       "3                     0.0                   0.0                   0.0   \n",
       "4                     0.0                   0.0                   0.0   \n",
       "\n",
       "   APPLICATION_TYPE_T3  APPLICATION_TYPE_T4  APPLICATION_TYPE_T5  \\\n",
       "0                  0.0                  0.0                  0.0   \n",
       "1                  1.0                  0.0                  0.0   \n",
       "2                  0.0                  0.0                  1.0   \n",
       "3                  1.0                  0.0                  0.0   \n",
       "4                  1.0                  0.0                  0.0   \n",
       "\n",
       "   APPLICATION_TYPE_T6  APPLICATION_TYPE_T7  APPLICATION_TYPE_T8  \\\n",
       "0                  0.0                  0.0                  0.0   \n",
       "1                  0.0                  0.0                  0.0   \n",
       "2                  0.0                  0.0                  0.0   \n",
       "3                  0.0                  0.0                  0.0   \n",
       "4                  0.0                  0.0                  0.0   \n",
       "\n",
       "   AFFILIATION_CompanySponsored  ...  INCOME_AMT_25000-99999  INCOME_AMT_50M+  \\\n",
       "0                           0.0  ...                     0.0              0.0   \n",
       "1                           0.0  ...                     0.0              0.0   \n",
       "2                           1.0  ...                     0.0              0.0   \n",
       "3                           1.0  ...                     0.0              0.0   \n",
       "4                           0.0  ...                     0.0              0.0   \n",
       "\n",
       "   INCOME_AMT_5M-10M  SPECIAL_CONSIDERATIONS_N  SPECIAL_CONSIDERATIONS_Y  \\\n",
       "0                0.0                       1.0                       0.0   \n",
       "1                0.0                       1.0                       0.0   \n",
       "2                0.0                       1.0                       0.0   \n",
       "3                0.0                       1.0                       0.0   \n",
       "4                0.0                       1.0                       0.0   \n",
       "\n",
       "   ASK_AMT_100000-999999  ASK_AMT_1000000-9999999  ASK_AMT_10000000+  \\\n",
       "0                    0.0                      0.0                0.0   \n",
       "1                    1.0                      0.0                0.0   \n",
       "2                    0.0                      0.0                0.0   \n",
       "3                    0.0                      0.0                0.0   \n",
       "4                    1.0                      0.0                0.0   \n",
       "\n",
       "   ASK_AMT_5000  ASK_AMT_5001-99999  \n",
       "0           1.0                 0.0  \n",
       "1           0.0                 0.0  \n",
       "2           1.0                 0.0  \n",
       "3           0.0                 1.0  \n",
       "4           0.0                 0.0  \n",
       "\n",
       "[5 rows x 46 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a OneHotEncoder instance\n",
    "enc = OneHotEncoder(sparse=False)\n",
    "\n",
    "# Fit and transform the OneHotEncoder using the categorical variable list\n",
    "encode_df = pd.DataFrame(enc.fit_transform(application_df[application_cat]))\n",
    "\n",
    "# Add the encoded variable names to the dataframe\n",
    "encode_df.columns = enc.get_feature_names_out(application_cat)\n",
    "encode_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>STATUS</th>\n",
       "      <th>IS_SUCCESSFUL</th>\n",
       "      <th>APPLICATION_TYPE_Other</th>\n",
       "      <th>APPLICATION_TYPE_T10</th>\n",
       "      <th>APPLICATION_TYPE_T19</th>\n",
       "      <th>APPLICATION_TYPE_T3</th>\n",
       "      <th>APPLICATION_TYPE_T4</th>\n",
       "      <th>APPLICATION_TYPE_T5</th>\n",
       "      <th>APPLICATION_TYPE_T6</th>\n",
       "      <th>APPLICATION_TYPE_T7</th>\n",
       "      <th>...</th>\n",
       "      <th>INCOME_AMT_25000-99999</th>\n",
       "      <th>INCOME_AMT_50M+</th>\n",
       "      <th>INCOME_AMT_5M-10M</th>\n",
       "      <th>SPECIAL_CONSIDERATIONS_N</th>\n",
       "      <th>SPECIAL_CONSIDERATIONS_Y</th>\n",
       "      <th>ASK_AMT_100000-999999</th>\n",
       "      <th>ASK_AMT_1000000-9999999</th>\n",
       "      <th>ASK_AMT_10000000+</th>\n",
       "      <th>ASK_AMT_5000</th>\n",
       "      <th>ASK_AMT_5001-99999</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</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>0.0</td>\n",
       "      <td>0.0</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</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>0.0</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>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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>0.0</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</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>0.0</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>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>1</td>\n",
       "      <td>1</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>0.0</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>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 48 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   STATUS  IS_SUCCESSFUL  APPLICATION_TYPE_Other  APPLICATION_TYPE_T10  \\\n",
       "0       1              1                     0.0                   1.0   \n",
       "1       1              1                     0.0                   0.0   \n",
       "2       1              0                     0.0                   0.0   \n",
       "3       1              1                     0.0                   0.0   \n",
       "4       1              1                     0.0                   0.0   \n",
       "\n",
       "   APPLICATION_TYPE_T19  APPLICATION_TYPE_T3  APPLICATION_TYPE_T4  \\\n",
       "0                   0.0                  0.0                  0.0   \n",
       "1                   0.0                  1.0                  0.0   \n",
       "2                   0.0                  0.0                  0.0   \n",
       "3                   0.0                  1.0                  0.0   \n",
       "4                   0.0                  1.0                  0.0   \n",
       "\n",
       "   APPLICATION_TYPE_T5  APPLICATION_TYPE_T6  APPLICATION_TYPE_T7  ...  \\\n",
       "0                  0.0                  0.0                  0.0  ...   \n",
       "1                  0.0                  0.0                  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",
       "   INCOME_AMT_25000-99999  INCOME_AMT_50M+  INCOME_AMT_5M-10M  \\\n",
       "0                     0.0              0.0                0.0   \n",
       "1                     0.0              0.0                0.0   \n",
       "2                     0.0              0.0                0.0   \n",
       "3                     0.0              0.0                0.0   \n",
       "4                     0.0              0.0                0.0   \n",
       "\n",
       "   SPECIAL_CONSIDERATIONS_N  SPECIAL_CONSIDERATIONS_Y  ASK_AMT_100000-999999  \\\n",
       "0                       1.0                       0.0                    0.0   \n",
       "1                       1.0                       0.0                    1.0   \n",
       "2                       1.0                       0.0                    0.0   \n",
       "3                       1.0                       0.0                    0.0   \n",
       "4                       1.0                       0.0                    1.0   \n",
       "\n",
       "   ASK_AMT_1000000-9999999  ASK_AMT_10000000+  ASK_AMT_5000  \\\n",
       "0                      0.0                0.0           1.0   \n",
       "1                      0.0                0.0           0.0   \n",
       "2                      0.0                0.0           1.0   \n",
       "3                      0.0                0.0           0.0   \n",
       "4                      0.0                0.0           0.0   \n",
       "\n",
       "   ASK_AMT_5001-99999  \n",
       "0                 0.0  \n",
       "1                 0.0  \n",
       "2                 0.0  \n",
       "3                 1.0  \n",
       "4                 0.0  \n",
       "\n",
       "[5 rows x 48 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Merge one-hot encoded features and drop the originals\n",
    "application_df = application_df.merge(encode_df, left_index=True, right_index=True)\n",
    "application_df = application_df.drop(application_cat, axis=1)\n",
    "application_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split our preprocessed data into our features and target arrays\n",
    "X = application_df.drop(\"IS_SUCCESSFUL\", axis=1).values\n",
    "y = application_df.IS_SUCCESSFUL.values\n",
    "\n",
    "# Split the preprocessed data into a training and testing dataset\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a StandardScaler instances\n",
    "scaler = StandardScaler()\n",
    "\n",
    "# Fit the StandardScaler\n",
    "X_scaler = scaler.fit(X_train)\n",
    "\n",
    "# Scale the data\n",
    "X_train_scaled = X_scaler.transform(X_train)\n",
    "X_test_scaled = X_scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### - 1st attempt : Additional neurons are added to hidden layers -"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense (Dense)               (None, 130)               6240      \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 90)                11790     \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 1)                 91        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 18,121\n",
      "Trainable params: 18,121\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Define the model - deep neural net, i.e., the number of input features and hidden nodes for each layer.\n",
    "number_input_features = len(X_train_scaled[0])\n",
    "hidden_nodes_layer1 = 130\n",
    "hidden_nodes_layer2 = 90\n",
    "\n",
    "nn = tf.keras.models.Sequential()\n",
    "\n",
    "# First hidden layer\n",
    "nn.add(tf.keras.layers.Dense(units=hidden_nodes_layer1, input_dim=number_input_features, activation=\"relu\"))\n",
    "\n",
    "# Second hidden layer\n",
    "nn.add(tf.keras.layers.Dense(units=hidden_nodes_layer2, activation=\"relu\"))\n",
    "\n",
    "# Output layer\n",
    "nn.add(tf.keras.layers.Dense(units=1, activation=\"sigmoid\"))\n",
    "\n",
    "# Check the structure of the model\n",
    "nn.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "nn.compile(loss=\"binary_crossentropy\", optimizer=\"adam\", metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:`period` argument is deprecated. Please use `save_freq` to specify the frequency in number of batches seen.\n",
      "Epoch 1/100\n",
      "804/804 [==============================] - 1s 725us/step - loss: 0.5680 - accuracy: 0.7266\n",
      "Epoch 2/100\n",
      "804/804 [==============================] - 1s 703us/step - loss: 0.5533 - accuracy: 0.7304\n",
      "Epoch 3/100\n",
      "804/804 [==============================] - 1s 706us/step - loss: 0.5502 - accuracy: 0.7323\n",
      "Epoch 4/100\n",
      "804/804 [==============================] - 1s 716us/step - loss: 0.5478 - accuracy: 0.7351\n",
      "Epoch 5/100\n",
      "753/804 [===========================>..] - ETA: 0s - loss: 0.5462 - accuracy: 0.7352\n",
      "Epoch 00005: saving model to checkpoints\\weights.05.hdf5\n",
      "804/804 [==============================] - 1s 751us/step - loss: 0.5462 - accuracy: 0.7350\n",
      "Epoch 6/100\n",
      "804/804 [==============================] - 1s 749us/step - loss: 0.5453 - accuracy: 0.73460s - loss: 0.5450 - accuracy: 0.73\n",
      "Epoch 7/100\n",
      "804/804 [==============================] - 1s 723us/step - loss: 0.5442 - accuracy: 0.7356\n",
      "Epoch 8/100\n",
      "804/804 [==============================] - 1s 706us/step - loss: 0.5440 - accuracy: 0.7367\n",
      "Epoch 9/100\n",
      "804/804 [==============================] - 1s 707us/step - loss: 0.5424 - accuracy: 0.7374\n",
      "Epoch 10/100\n",
      "787/804 [============================>.] - ETA: 0s - loss: 0.5413 - accuracy: 0.7384\n",
      "Epoch 00010: saving model to checkpoints\\weights.10.hdf5\n",
      "804/804 [==============================] - 1s 723us/step - loss: 0.5415 - accuracy: 0.7386\n",
      "Epoch 11/100\n",
      "804/804 [==============================] - 1s 729us/step - loss: 0.5411 - accuracy: 0.7381\n",
      "Epoch 12/100\n",
      "804/804 [==============================] - 1s 708us/step - loss: 0.5411 - accuracy: 0.7388\n",
      "Epoch 13/100\n",
      "804/804 [==============================] - 1s 732us/step - loss: 0.5401 - accuracy: 0.7399\n",
      "Epoch 14/100\n",
      "804/804 [==============================] - 1s 715us/step - loss: 0.5403 - accuracy: 0.7385\n",
      "Epoch 15/100\n",
      "803/804 [============================>.] - ETA: 0s - loss: 0.5385 - accuracy: 0.7393\n",
      "Epoch 00015: saving model to checkpoints\\weights.15.hdf5\n",
      "804/804 [==============================] - 1s 709us/step - loss: 0.5385 - accuracy: 0.7393\n",
      "Epoch 16/100\n",
      "804/804 [==============================] - 1s 727us/step - loss: 0.5388 - accuracy: 0.7405\n",
      "Epoch 17/100\n",
      "804/804 [==============================] - 1s 711us/step - loss: 0.5372 - accuracy: 0.7411\n",
      "Epoch 18/100\n",
      "804/804 [==============================] - 1s 734us/step - loss: 0.5380 - accuracy: 0.7403\n",
      "Epoch 19/100\n",
      "804/804 [==============================] - 1s 700us/step - loss: 0.5374 - accuracy: 0.7411\n",
      "Epoch 20/100\n",
      "779/804 [============================>.] - ETA: 0s - loss: 0.5371 - accuracy: 0.7399\n",
      "Epoch 00020: saving model to checkpoints\\weights.20.hdf5\n",
      "804/804 [==============================] - 1s 731us/step - loss: 0.5365 - accuracy: 0.7402\n",
      "Epoch 21/100\n",
      "804/804 [==============================] - 1s 710us/step - loss: 0.5363 - accuracy: 0.7408\n",
      "Epoch 22/100\n",
      "804/804 [==============================] - 1s 728us/step - loss: 0.5361 - accuracy: 0.7408\n",
      "Epoch 23/100\n",
      "804/804 [==============================] - 1s 705us/step - loss: 0.5357 - accuracy: 0.7413\n",
      "Epoch 24/100\n",
      "804/804 [==============================] - 1s 703us/step - loss: 0.5351 - accuracy: 0.7425\n",
      "Epoch 25/100\n",
      "790/804 [============================>.] - ETA: 0s - loss: 0.5346 - accuracy: 0.7414\n",
      "Epoch 00025: saving model to checkpoints\\weights.25.hdf5\n",
      "804/804 [==============================] - 1s 726us/step - loss: 0.5350 - accuracy: 0.7413\n",
      "Epoch 26/100\n",
      "804/804 [==============================] - 1s 710us/step - loss: 0.5347 - accuracy: 0.7416\n",
      "Epoch 27/100\n",
      "804/804 [==============================] - 1s 724us/step - loss: 0.5339 - accuracy: 0.7411\n",
      "Epoch 28/100\n",
      "804/804 [==============================] - 1s 709us/step - loss: 0.5334 - accuracy: 0.7436\n",
      "Epoch 29/100\n",
      "804/804 [==============================] - 1s 701us/step - loss: 0.5334 - accuracy: 0.7425\n",
      "Epoch 30/100\n",
      "792/804 [============================>.] - ETA: 0s - loss: 0.5331 - accuracy: 0.7412\n",
      "Epoch 00030: saving model to checkpoints\\weights.30.hdf5\n",
      "804/804 [==============================] - 1s 719us/step - loss: 0.5335 - accuracy: 0.7409\n",
      "Epoch 31/100\n",
      "804/804 [==============================] - 1s 711us/step - loss: 0.5326 - accuracy: 0.7432\n",
      "Epoch 32/100\n",
      "804/804 [==============================] - 1s 710us/step - loss: 0.5326 - accuracy: 0.7434\n",
      "Epoch 33/100\n",
      "804/804 [==============================] - 1s 718us/step - loss: 0.5320 - accuracy: 0.7432\n",
      "Epoch 34/100\n",
      "804/804 [==============================] - 1s 715us/step - loss: 0.5321 - accuracy: 0.7434\n",
      "Epoch 35/100\n",
      "787/804 [============================>.] - ETA: 0s - loss: 0.5319 - accuracy: 0.7428\n",
      "Epoch 00035: saving model to checkpoints\\weights.35.hdf5\n",
      "804/804 [==============================] - 1s 722us/step - loss: 0.5323 - accuracy: 0.7428\n",
      "Epoch 36/100\n",
      "804/804 [==============================] - 1s 729us/step - loss: 0.5309 - accuracy: 0.7444\n",
      "Epoch 37/100\n",
      "804/804 [==============================] - 1s 702us/step - loss: 0.5317 - accuracy: 0.7431\n",
      "Epoch 38/100\n",
      "804/804 [==============================] - 1s 733us/step - loss: 0.5310 - accuracy: 0.7436\n",
      "Epoch 39/100\n",
      "804/804 [==============================] - 1s 731us/step - loss: 0.5308 - accuracy: 0.7453\n",
      "Epoch 40/100\n",
      "786/804 [============================>.] - ETA: 0s - loss: 0.5296 - accuracy: 0.7454\n",
      "Epoch 00040: saving model to checkpoints\\weights.40.hdf5\n",
      "804/804 [==============================] - 1s 724us/step - loss: 0.5301 - accuracy: 0.7446\n",
      "Epoch 41/100\n",
      "804/804 [==============================] - 1s 749us/step - loss: 0.5307 - accuracy: 0.7439\n",
      "Epoch 42/100\n",
      "804/804 [==============================] - 1s 706us/step - loss: 0.5300 - accuracy: 0.7444\n",
      "Epoch 43/100\n",
      "804/804 [==============================] - 1s 708us/step - loss: 0.5298 - accuracy: 0.7451\n",
      "Epoch 44/100\n",
      "804/804 [==============================] - 1s 708us/step - loss: 0.5300 - accuracy: 0.7451\n",
      "Epoch 45/100\n",
      "785/804 [============================>.] - ETA: 0s - loss: 0.5290 - accuracy: 0.7461\n",
      "Epoch 00045: saving model to checkpoints\\weights.45.hdf5\n",
      "804/804 [==============================] - 1s 721us/step - loss: 0.5300 - accuracy: 0.7455\n",
      "Epoch 46/100\n",
      "804/804 [==============================] - 1s 705us/step - loss: 0.5292 - accuracy: 0.7447\n",
      "Epoch 47/100\n",
      "804/804 [==============================] - 1s 709us/step - loss: 0.5295 - accuracy: 0.7442\n",
      "Epoch 48/100\n",
      "804/804 [==============================] - 1s 718us/step - loss: 0.5293 - accuracy: 0.7452\n",
      "Epoch 49/100\n",
      "804/804 [==============================] - 1s 702us/step - loss: 0.5293 - accuracy: 0.7449\n",
      "Epoch 50/100\n",
      "780/804 [============================>.] - ETA: 0s - loss: 0.5291 - accuracy: 0.7454\n",
      "Epoch 00050: saving model to checkpoints\\weights.50.hdf5\n",
      "804/804 [==============================] - 1s 727us/step - loss: 0.5291 - accuracy: 0.7455\n",
      "Epoch 51/100\n",
      "804/804 [==============================] - 1s 717us/step - loss: 0.5282 - accuracy: 0.74460s - loss: 0.5265 - accuracy\n",
      "Epoch 52/100\n",
      "804/804 [==============================] - 1s 707us/step - loss: 0.5286 - accuracy: 0.7446\n",
      "Epoch 53/100\n",
      "804/804 [==============================] - 1s 728us/step - loss: 0.5288 - accuracy: 0.7440\n",
      "Epoch 54/100\n",
      "804/804 [==============================] - 1s 707us/step - loss: 0.5281 - accuracy: 0.7451\n",
      "Epoch 55/100\n",
      "784/804 [============================>.] - ETA: 0s - loss: 0.5281 - accuracy: 0.7456\n",
      "Epoch 00055: saving model to checkpoints\\weights.55.hdf5\n",
      "804/804 [==============================] - 1s 725us/step - loss: 0.5280 - accuracy: 0.7456\n",
      "Epoch 56/100\n",
      "804/804 [==============================] - 1s 728us/step - loss: 0.5282 - accuracy: 0.7448\n",
      "Epoch 57/100\n",
      "804/804 [==============================] - 1s 710us/step - loss: 0.5275 - accuracy: 0.7462\n",
      "Epoch 58/100\n",
      "804/804 [==============================] - 1s 707us/step - loss: 0.5279 - accuracy: 0.7447\n",
      "Epoch 59/100\n",
      "804/804 [==============================] - 1s 708us/step - loss: 0.5281 - accuracy: 0.7454\n",
      "Epoch 60/100\n",
      "792/804 [============================>.] - ETA: 0s - loss: 0.5274 - accuracy: 0.7450\n",
      "Epoch 00060: saving model to checkpoints\\weights.60.hdf5\n",
      "804/804 [==============================] - 1s 716us/step - loss: 0.5274 - accuracy: 0.7449\n",
      "Epoch 61/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "804/804 [==============================] - 1s 705us/step - loss: 0.5268 - accuracy: 0.7456\n",
      "Epoch 62/100\n",
      "804/804 [==============================] - 1s 698us/step - loss: 0.5272 - accuracy: 0.7443\n",
      "Epoch 63/100\n",
      "804/804 [==============================] - 1s 693us/step - loss: 0.5270 - accuracy: 0.7448\n",
      "Epoch 64/100\n",
      "804/804 [==============================] - 1s 719us/step - loss: 0.5270 - accuracy: 0.7454\n",
      "Epoch 65/100\n",
      "755/804 [===========================>..] - ETA: 0s - loss: 0.5267 - accuracy: 0.7464\n",
      "Epoch 00065: saving model to checkpoints\\weights.65.hdf5\n",
      "804/804 [==============================] - 1s 756us/step - loss: 0.5267 - accuracy: 0.7468\n",
      "Epoch 66/100\n",
      "804/804 [==============================] - 1s 730us/step - loss: 0.5268 - accuracy: 0.7456\n",
      "Epoch 67/100\n",
      "804/804 [==============================] - 1s 731us/step - loss: 0.5272 - accuracy: 0.7456\n",
      "Epoch 68/100\n",
      "804/804 [==============================] - 1s 719us/step - loss: 0.5265 - accuracy: 0.7460\n",
      "Epoch 69/100\n",
      "804/804 [==============================] - 1s 750us/step - loss: 0.5267 - accuracy: 0.7466\n",
      "Epoch 70/100\n",
      "770/804 [===========================>..] - ETA: 0s - loss: 0.5254 - accuracy: 0.7465\n",
      "Epoch 00070: saving model to checkpoints\\weights.70.hdf5\n",
      "804/804 [==============================] - 1s 738us/step - loss: 0.5263 - accuracy: 0.7457\n",
      "Epoch 71/100\n",
      "804/804 [==============================] - 1s 730us/step - loss: 0.5265 - accuracy: 0.7451\n",
      "Epoch 72/100\n",
      "804/804 [==============================] - 1s 748us/step - loss: 0.5264 - accuracy: 0.7455\n",
      "Epoch 73/100\n",
      "804/804 [==============================] - 1s 734us/step - loss: 0.5259 - accuracy: 0.7449\n",
      "Epoch 74/100\n",
      "804/804 [==============================] - 1s 765us/step - loss: 0.5260 - accuracy: 0.7443\n",
      "Epoch 75/100\n",
      "781/804 [============================>.] - ETA: 0s - loss: 0.5255 - accuracy: 0.7455\n",
      "Epoch 00075: saving model to checkpoints\\weights.75.hdf5\n",
      "804/804 [==============================] - 1s 725us/step - loss: 0.5259 - accuracy: 0.7453\n",
      "Epoch 76/100\n",
      "804/804 [==============================] - 1s 727us/step - loss: 0.5260 - accuracy: 0.7452\n",
      "Epoch 77/100\n",
      "804/804 [==============================] - 1s 704us/step - loss: 0.5254 - accuracy: 0.7455\n",
      "Epoch 78/100\n",
      "804/804 [==============================] - 1s 702us/step - loss: 0.5258 - accuracy: 0.7462\n",
      "Epoch 79/100\n",
      "804/804 [==============================] - 1s 742us/step - loss: 0.5257 - accuracy: 0.7465\n",
      "Epoch 80/100\n",
      "759/804 [===========================>..] - ETA: 0s - loss: 0.5284 - accuracy: 0.7453\n",
      "Epoch 00080: saving model to checkpoints\\weights.80.hdf5\n",
      "804/804 [==============================] - 1s 745us/step - loss: 0.5279 - accuracy: 0.7462\n",
      "Epoch 81/100\n",
      "804/804 [==============================] - 1s 715us/step - loss: 0.5256 - accuracy: 0.7460\n",
      "Epoch 82/100\n",
      "804/804 [==============================] - 1s 758us/step - loss: 0.5251 - accuracy: 0.7463\n",
      "Epoch 83/100\n",
      "804/804 [==============================] - 1s 712us/step - loss: 0.5257 - accuracy: 0.7465\n",
      "Epoch 84/100\n",
      "804/804 [==============================] - 1s 758us/step - loss: 0.5262 - accuracy: 0.7462\n",
      "Epoch 85/100\n",
      "793/804 [============================>.] - ETA: 0s - loss: 0.5239 - accuracy: 0.7472 ETA: 0s - loss: 0.5222 - accu\n",
      "Epoch 00085: saving model to checkpoints\\weights.85.hdf5\n",
      "804/804 [==============================] - 1s 716us/step - loss: 0.5248 - accuracy: 0.7464\n",
      "Epoch 86/100\n",
      "804/804 [==============================] - 1s 714us/step - loss: 0.5248 - accuracy: 0.7460\n",
      "Epoch 87/100\n",
      "804/804 [==============================] - 1s 708us/step - loss: 0.5252 - accuracy: 0.7465\n",
      "Epoch 88/100\n",
      "804/804 [==============================] - 1s 711us/step - loss: 0.5252 - accuracy: 0.7462\n",
      "Epoch 89/100\n",
      "804/804 [==============================] - 1s 710us/step - loss: 0.5248 - accuracy: 0.7461\n",
      "Epoch 90/100\n",
      "784/804 [============================>.] - ETA: 0s - loss: 0.5248 - accuracy: 0.7455\n",
      "Epoch 00090: saving model to checkpoints\\weights.90.hdf5\n",
      "804/804 [==============================] - 1s 726us/step - loss: 0.5246 - accuracy: 0.7458\n",
      "Epoch 91/100\n",
      "804/804 [==============================] - 1s 741us/step - loss: 0.5244 - accuracy: 0.7457\n",
      "Epoch 92/100\n",
      "804/804 [==============================] - 1s 708us/step - loss: 0.5251 - accuracy: 0.7461\n",
      "Epoch 93/100\n",
      "804/804 [==============================] - 1s 710us/step - loss: 0.5245 - accuracy: 0.7462\n",
      "Epoch 94/100\n",
      "804/804 [==============================] - 1s 743us/step - loss: 0.5254 - accuracy: 0.7462\n",
      "Epoch 95/100\n",
      "745/804 [==========================>...] - ETA: 0s - loss: 0.5246 - accuracy: 0.7466\n",
      "Epoch 00095: saving model to checkpoints\\weights.95.hdf5\n",
      "804/804 [==============================] - 1s 763us/step - loss: 0.5244 - accuracy: 0.7464\n",
      "Epoch 96/100\n",
      "804/804 [==============================] - 1s 707us/step - loss: 0.5250 - accuracy: 0.7460\n",
      "Epoch 97/100\n",
      "804/804 [==============================] - 1s 698us/step - loss: 0.5287 - accuracy: 0.7464\n",
      "Epoch 98/100\n",
      "804/804 [==============================] - 1s 739us/step - loss: 0.5270 - accuracy: 0.7465\n",
      "Epoch 99/100\n",
      "804/804 [==============================] - 1s 720us/step - loss: 0.5235 - accuracy: 0.7470\n",
      "Epoch 100/100\n",
      "789/804 [============================>.] - ETA: 0s - loss: 0.5241 - accuracy: 0.7459\n",
      "Epoch 00100: saving model to checkpoints\\weights.100.hdf5\n",
      "804/804 [==============================] - 1s 720us/step - loss: 0.5243 - accuracy: 0.7456\n"
     ]
    }
   ],
   "source": [
    "# Define the checkpoint path and filenames\n",
    "os.makedirs(\"checkpoints_opt/\", exist_ok=True)\n",
    "checkpoint_path = \"checkpoints/weights.{epoch:02d}.hdf5\"\n",
    "\n",
    "# Create a callback that saves the model's weights every epoch\n",
    "cp_callback = ModelCheckpoint(\n",
    "    filepath=checkpoint_path,\n",
    "    verbose=1,\n",
    "    save_weights_only=True,\n",
    "    save_freq=\"epoch\",\n",
    "    period=5)\n",
    "\n",
    "# Train the model\n",
    "fit_model = nn.fit(X_train_scaled, y_train, epochs=100, callbacks=[cp_callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "268/268 - 0s - loss: 0.5906 - accuracy: 0.7235 - 213ms/epoch - 796us/step\n",
      "Loss: 0.5905671715736389, Accuracy: 0.723498523235321\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model using the test data\n",
    "model_loss, model_accuracy = nn.evaluate(X_test_scaled, y_test, verbose=2)\n",
    "print(f\"Loss: {model_loss}, Accuracy: {model_accuracy}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Export our model to HDF5 file\n",
    "nn.save(\"AlphabetSoupCharity_Optimization.h5\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### - 2nd attempt : Additional hidden layers are added -"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_3 (Dense)             (None, 130)               6240      \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 90)                11790     \n",
      "                                                                 \n",
      " dense_5 (Dense)             (None, 30)                2730      \n",
      "                                                                 \n",
      " dense_6 (Dense)             (None, 1)                 31        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 20,791\n",
      "Trainable params: 20,791\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Define the model - deep neural net, i.e., the number of input features and hidden nodes for each layer.\n",
    "number_input_features = len(X_train_scaled[0])\n",
    "hidden_nodes_layer1 = 130\n",
    "hidden_nodes_layer2 = 90\n",
    "hidden_nodes_layer3 = 30\n",
    "\n",
    "nn = tf.keras.models.Sequential()\n",
    "\n",
    "# First hidden layer\n",
    "nn.add(tf.keras.layers.Dense(units=hidden_nodes_layer1, input_dim=number_input_features, activation=\"relu\"))\n",
    "\n",
    "# Second hidden layer\n",
    "nn.add(tf.keras.layers.Dense(units=hidden_nodes_layer2, activation=\"relu\"))\n",
    "\n",
    "# Third hidden layer\n",
    "nn.add(tf.keras.layers.Dense(units=hidden_nodes_layer3, activation=\"relu\"))\n",
    "\n",
    "# Output layer\n",
    "nn.add(tf.keras.layers.Dense(units=1, activation=\"sigmoid\"))\n",
    "\n",
    "# Check the structure of the model\n",
    "nn.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "nn.compile(loss=\"binary_crossentropy\", optimizer=\"adam\", metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:`period` argument is deprecated. Please use `save_freq` to specify the frequency in number of batches seen.\n",
      "Epoch 1/100\n",
      "804/804 [==============================] - 1s 776us/step - loss: 0.5680 - accuracy: 0.7233\n",
      "Epoch 2/100\n",
      "804/804 [==============================] - 1s 756us/step - loss: 0.5524 - accuracy: 0.7327\n",
      "Epoch 3/100\n",
      "804/804 [==============================] - 1s 750us/step - loss: 0.5495 - accuracy: 0.7324\n",
      "Epoch 4/100\n",
      "804/804 [==============================] - 1s 772us/step - loss: 0.5471 - accuracy: 0.7337\n",
      "Epoch 5/100\n",
      "792/804 [============================>.] - ETA: 0s - loss: 0.5467 - accuracy: 0.7351\n",
      "Epoch 00005: saving model to checkpoints\\weights.05.hdf5\n",
      "804/804 [==============================] - 1s 787us/step - loss: 0.5460 - accuracy: 0.7352\n",
      "Epoch 6/100\n",
      "804/804 [==============================] - 1s 741us/step - loss: 0.5441 - accuracy: 0.7349\n",
      "Epoch 7/100\n",
      "804/804 [==============================] - 1s 746us/step - loss: 0.5440 - accuracy: 0.7351\n",
      "Epoch 8/100\n",
      "804/804 [==============================] - 1s 790us/step - loss: 0.5430 - accuracy: 0.7378\n",
      "Epoch 9/100\n",
      "804/804 [==============================] - 1s 758us/step - loss: 0.5425 - accuracy: 0.7369\n",
      "Epoch 10/100\n",
      "742/804 [==========================>...] - ETA: 0s - loss: 0.5409 - accuracy: 0.7391\n",
      "Epoch 00010: saving model to checkpoints\\weights.10.hdf5\n",
      "804/804 [==============================] - 1s 765us/step - loss: 0.5416 - accuracy: 0.7386\n",
      "Epoch 11/100\n",
      "804/804 [==============================] - 1s 815us/step - loss: 0.5411 - accuracy: 0.7381\n",
      "Epoch 12/100\n",
      "804/804 [==============================] - 1s 803us/step - loss: 0.5405 - accuracy: 0.7395\n",
      "Epoch 13/100\n",
      "804/804 [==============================] - 1s 786us/step - loss: 0.5397 - accuracy: 0.7383\n",
      "Epoch 14/100\n",
      "804/804 [==============================] - 1s 789us/step - loss: 0.5398 - accuracy: 0.7383\n",
      "Epoch 15/100\n",
      "791/804 [============================>.] - ETA: 0s - loss: 0.5385 - accuracy: 0.7400 ETA: 0s - loss: 0.5389 \n",
      "Epoch 00015: saving model to checkpoints\\weights.15.hdf5\n",
      "804/804 [==============================] - 1s 783us/step - loss: 0.5388 - accuracy: 0.7396\n",
      "Epoch 16/100\n",
      "804/804 [==============================] - 1s 806us/step - loss: 0.5386 - accuracy: 0.7390\n",
      "Epoch 17/100\n",
      "804/804 [==============================] - 1s 775us/step - loss: 0.5372 - accuracy: 0.7407\n",
      "Epoch 18/100\n",
      "804/804 [==============================] - 1s 799us/step - loss: 0.5367 - accuracy: 0.7401\n",
      "Epoch 19/100\n",
      "804/804 [==============================] - 1s 796us/step - loss: 0.5366 - accuracy: 0.7416\n",
      "Epoch 20/100\n",
      "787/804 [============================>.] - ETA: 0s - loss: 0.5358 - accuracy: 0.7414\n",
      "Epoch 00020: saving model to checkpoints\\weights.20.hdf5\n",
      "804/804 [==============================] - 1s 786us/step - loss: 0.5357 - accuracy: 0.7415\n",
      "Epoch 21/100\n",
      "804/804 [==============================] - 1s 763us/step - loss: 0.5358 - accuracy: 0.7420\n",
      "Epoch 22/100\n",
      "804/804 [==============================] - 1s 768us/step - loss: 0.5349 - accuracy: 0.7413\n",
      "Epoch 23/100\n",
      "804/804 [==============================] - 1s 768us/step - loss: 0.5349 - accuracy: 0.7420\n",
      "Epoch 24/100\n",
      "804/804 [==============================] - 1s 749us/step - loss: 0.5343 - accuracy: 0.7414\n",
      "Epoch 25/100\n",
      "786/804 [============================>.] - ETA: 0s - loss: 0.5332 - accuracy: 0.7429\n",
      "Epoch 00025: saving model to checkpoints\\weights.25.hdf5\n",
      "804/804 [==============================] - 1s 789us/step - loss: 0.5341 - accuracy: 0.7424\n",
      "Epoch 26/100\n",
      "804/804 [==============================] - 1s 730us/step - loss: 0.5330 - accuracy: 0.7423\n",
      "Epoch 27/100\n",
      "804/804 [==============================] - 1s 742us/step - loss: 0.5335 - accuracy: 0.7423\n",
      "Epoch 28/100\n",
      "804/804 [==============================] - 1s 834us/step - loss: 0.5319 - accuracy: 0.7428\n",
      "Epoch 29/100\n",
      "804/804 [==============================] - 1s 769us/step - loss: 0.5323 - accuracy: 0.7431\n",
      "Epoch 30/100\n",
      "783/804 [============================>.] - ETA: 0s - loss: 0.5321 - accuracy: 0.7430\n",
      "Epoch 00030: saving model to checkpoints\\weights.30.hdf5\n",
      "804/804 [==============================] - 1s 793us/step - loss: 0.5314 - accuracy: 0.7437\n",
      "Epoch 31/100\n",
      "804/804 [==============================] - 1s 746us/step - loss: 0.5315 - accuracy: 0.7438\n",
      "Epoch 32/100\n",
      "804/804 [==============================] - 1s 732us/step - loss: 0.5309 - accuracy: 0.7439\n",
      "Epoch 33/100\n",
      "804/804 [==============================] - 1s 798us/step - loss: 0.5309 - accuracy: 0.7439\n",
      "Epoch 34/100\n",
      "804/804 [==============================] - 1s 752us/step - loss: 0.5308 - accuracy: 0.74460s - loss: 0.5\n",
      "Epoch 35/100\n",
      "795/804 [============================>.] - ETA: 0s - loss: 0.5298 - accuracy: 0.7429\n",
      "Epoch 00035: saving model to checkpoints\\weights.35.hdf5\n",
      "804/804 [==============================] - 1s 784us/step - loss: 0.5302 - accuracy: 0.7423\n",
      "Epoch 36/100\n",
      "804/804 [==============================] - 1s 750us/step - loss: 0.5302 - accuracy: 0.7435\n",
      "Epoch 37/100\n",
      "804/804 [==============================] - 1s 810us/step - loss: 0.5291 - accuracy: 0.7445\n",
      "Epoch 38/100\n",
      "804/804 [==============================] - 1s 832us/step - loss: 0.5296 - accuracy: 0.7442\n",
      "Epoch 39/100\n",
      "804/804 [==============================] - 1s 733us/step - loss: 0.5294 - accuracy: 0.7437\n",
      "Epoch 40/100\n",
      "767/804 [===========================>..] - ETA: 0s - loss: 0.5288 - accuracy: 0.7450\n",
      "Epoch 00040: saving model to checkpoints\\weights.40.hdf5\n",
      "804/804 [==============================] - 1s 808us/step - loss: 0.5287 - accuracy: 0.7449\n",
      "Epoch 41/100\n",
      "804/804 [==============================] - 1s 775us/step - loss: 0.5282 - accuracy: 0.7448\n",
      "Epoch 42/100\n",
      "804/804 [==============================] - 1s 746us/step - loss: 0.5282 - accuracy: 0.7442\n",
      "Epoch 43/100\n",
      "804/804 [==============================] - 1s 788us/step - loss: 0.5281 - accuracy: 0.7451\n",
      "Epoch 44/100\n",
      "804/804 [==============================] - 1s 763us/step - loss: 0.5277 - accuracy: 0.7435\n",
      "Epoch 45/100\n",
      "746/804 [==========================>...] - ETA: 0s - loss: 0.5274 - accuracy: 0.7457\n",
      "Epoch 00045: saving model to checkpoints\\weights.45.hdf5\n",
      "804/804 [==============================] - 1s 760us/step - loss: 0.5273 - accuracy: 0.7451\n",
      "Epoch 46/100\n",
      "804/804 [==============================] - 1s 751us/step - loss: 0.5278 - accuracy: 0.7455\n",
      "Epoch 47/100\n",
      "804/804 [==============================] - 1s 802us/step - loss: 0.5267 - accuracy: 0.7449\n",
      "Epoch 48/100\n",
      "804/804 [==============================] - 1s 763us/step - loss: 0.5271 - accuracy: 0.7448\n",
      "Epoch 49/100\n",
      "804/804 [==============================] - 1s 753us/step - loss: 0.5268 - accuracy: 0.7446\n",
      "Epoch 50/100\n",
      "744/804 [==========================>...] - ETA: 0s - loss: 0.5252 - accuracy: 0.7476\n",
      "Epoch 00050: saving model to checkpoints\\weights.50.hdf5\n",
      "804/804 [==============================] - 1s 762us/step - loss: 0.5263 - accuracy: 0.7462\n",
      "Epoch 51/100\n",
      "804/804 [==============================] - 1s 746us/step - loss: 0.5266 - accuracy: 0.7446\n",
      "Epoch 52/100\n",
      "804/804 [==============================] - 1s 767us/step - loss: 0.5264 - accuracy: 0.7462\n",
      "Epoch 53/100\n",
      "804/804 [==============================] - 1s 734us/step - loss: 0.5262 - accuracy: 0.7454\n",
      "Epoch 54/100\n",
      "804/804 [==============================] - 1s 754us/step - loss: 0.5263 - accuracy: 0.7464\n",
      "Epoch 55/100\n",
      "746/804 [==========================>...] - ETA: 0s - loss: 0.5252 - accuracy: 0.7459\n",
      "Epoch 00055: saving model to checkpoints\\weights.55.hdf5\n",
      "804/804 [==============================] - 1s 826us/step - loss: 0.5258 - accuracy: 0.7455\n",
      "Epoch 56/100\n",
      "804/804 [==============================] - 1s 793us/step - loss: 0.5252 - accuracy: 0.7455\n",
      "Epoch 57/100\n",
      "804/804 [==============================] - 1s 763us/step - loss: 0.5253 - accuracy: 0.7454\n",
      "Epoch 58/100\n",
      "804/804 [==============================] - 1s 768us/step - loss: 0.5248 - accuracy: 0.7452\n",
      "Epoch 59/100\n",
      "804/804 [==============================] - 1s 754us/step - loss: 0.5247 - accuracy: 0.7460\n",
      "Epoch 60/100\n",
      "789/804 [============================>.] - ETA: 0s - loss: 0.5262 - accuracy: 0.7461 ETA: 0s - loss: 0.5236 - accuracy:  - ETA: 0s - loss: 0.5189 - accu\n",
      "Epoch 00060: saving model to checkpoints\\weights.60.hdf5\n",
      "804/804 [==============================] - 1s 787us/step - loss: 0.5264 - accuracy: 0.7461\n",
      "Epoch 61/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "804/804 [==============================] - 1s 811us/step - loss: 0.5254 - accuracy: 0.7458\n",
      "Epoch 62/100\n",
      "804/804 [==============================] - 1s 743us/step - loss: 0.5252 - accuracy: 0.7453\n",
      "Epoch 63/100\n",
      "804/804 [==============================] - 1s 753us/step - loss: 0.5267 - accuracy: 0.7448\n",
      "Epoch 64/100\n",
      "804/804 [==============================] - 1s 728us/step - loss: 0.5252 - accuracy: 0.7456\n",
      "Epoch 65/100\n",
      "766/804 [===========================>..] - ETA: 0s - loss: 0.5245 - accuracy: 0.7459\n",
      "Epoch 00065: saving model to checkpoints\\weights.65.hdf5\n",
      "804/804 [==============================] - 1s 742us/step - loss: 0.5245 - accuracy: 0.7460\n",
      "Epoch 66/100\n",
      "804/804 [==============================] - 1s 754us/step - loss: 0.5242 - accuracy: 0.7465\n",
      "Epoch 67/100\n",
      "804/804 [==============================] - 1s 780us/step - loss: 0.5242 - accuracy: 0.7463\n",
      "Epoch 68/100\n",
      "804/804 [==============================] - 1s 731us/step - loss: 0.5245 - accuracy: 0.7462\n",
      "Epoch 69/100\n",
      "804/804 [==============================] - 1s 733us/step - loss: 0.5253 - accuracy: 0.7470\n",
      "Epoch 70/100\n",
      "739/804 [==========================>...] - ETA: 0s - loss: 0.5248 - accuracy: 0.7462\n",
      "Epoch 00070: saving model to checkpoints\\weights.70.hdf5\n",
      "804/804 [==============================] - 1s 766us/step - loss: 0.5242 - accuracy: 0.7463\n",
      "Epoch 71/100\n",
      "804/804 [==============================] - 1s 814us/step - loss: 0.5235 - accuracy: 0.74700s - loss: 0.5175 - \n",
      "Epoch 72/100\n",
      "804/804 [==============================] - 1s 804us/step - loss: 0.5239 - accuracy: 0.7459\n",
      "Epoch 73/100\n",
      "804/804 [==============================] - 1s 737us/step - loss: 0.5240 - accuracy: 0.7469\n",
      "Epoch 74/100\n",
      "804/804 [==============================] - 1s 735us/step - loss: 0.5240 - accuracy: 0.7466\n",
      "Epoch 75/100\n",
      "760/804 [===========================>..] - ETA: 0s - loss: 0.5240 - accuracy: 0.7468\n",
      "Epoch 00075: saving model to checkpoints\\weights.75.hdf5\n",
      "804/804 [==============================] - 1s 749us/step - loss: 0.5243 - accuracy: 0.7468\n",
      "Epoch 76/100\n",
      "804/804 [==============================] - 1s 754us/step - loss: 0.5236 - accuracy: 0.7476\n",
      "Epoch 77/100\n",
      "804/804 [==============================] - 1s 766us/step - loss: 0.5244 - accuracy: 0.7464\n",
      "Epoch 78/100\n",
      "804/804 [==============================] - 1s 777us/step - loss: 0.5237 - accuracy: 0.7460\n",
      "Epoch 79/100\n",
      "804/804 [==============================] - 1s 765us/step - loss: 0.5243 - accuracy: 0.7463\n",
      "Epoch 80/100\n",
      "799/804 [============================>.] - ETA: 0s - loss: 0.5223 - accuracy: 0.7472\n",
      "Epoch 00080: saving model to checkpoints\\weights.80.hdf5\n",
      "804/804 [==============================] - 1s 776us/step - loss: 0.5228 - accuracy: 0.7467\n",
      "Epoch 81/100\n",
      "804/804 [==============================] - 1s 746us/step - loss: 0.5232 - accuracy: 0.7462\n",
      "Epoch 82/100\n",
      "804/804 [==============================] - 1s 810us/step - loss: 0.5237 - accuracy: 0.7461\n",
      "Epoch 83/100\n",
      "804/804 [==============================] - 1s 764us/step - loss: 0.5229 - accuracy: 0.7467\n",
      "Epoch 84/100\n",
      "804/804 [==============================] - 1s 735us/step - loss: 0.5234 - accuracy: 0.7467\n",
      "Epoch 85/100\n",
      "767/804 [===========================>..] - ETA: 0s - loss: 0.5220 - accuracy: 0.7475\n",
      "Epoch 00085: saving model to checkpoints\\weights.85.hdf5\n",
      "804/804 [==============================] - 1s 807us/step - loss: 0.5229 - accuracy: 0.7468\n",
      "Epoch 86/100\n",
      "804/804 [==============================] - 1s 754us/step - loss: 0.5225 - accuracy: 0.7470\n",
      "Epoch 87/100\n",
      "804/804 [==============================] - 1s 790us/step - loss: 0.5233 - accuracy: 0.7459\n",
      "Epoch 88/100\n",
      "804/804 [==============================] - 1s 769us/step - loss: 0.5229 - accuracy: 0.7458\n",
      "Epoch 89/100\n",
      "804/804 [==============================] - 1s 807us/step - loss: 0.5233 - accuracy: 0.7466\n",
      "Epoch 90/100\n",
      "756/804 [===========================>..] - ETA: 0s - loss: 0.5222 - accuracy: 0.7470\n",
      "Epoch 00090: saving model to checkpoints\\weights.90.hdf5\n",
      "804/804 [==============================] - 1s 816us/step - loss: 0.5225 - accuracy: 0.7470\n",
      "Epoch 91/100\n",
      "804/804 [==============================] - 1s 813us/step - loss: 0.5229 - accuracy: 0.7459\n",
      "Epoch 92/100\n",
      "804/804 [==============================] - 1s 771us/step - loss: 0.5234 - accuracy: 0.7460\n",
      "Epoch 93/100\n",
      "804/804 [==============================] - 1s 801us/step - loss: 0.5227 - accuracy: 0.7468\n",
      "Epoch 94/100\n",
      "804/804 [==============================] - 1s 791us/step - loss: 0.5245 - accuracy: 0.7468\n",
      "Epoch 95/100\n",
      "801/804 [============================>.] - ETA: 0s - loss: 0.5224 - accuracy: 0.7469\n",
      "Epoch 00095: saving model to checkpoints\\weights.95.hdf5\n",
      "804/804 [==============================] - 1s 774us/step - loss: 0.5222 - accuracy: 0.7470\n",
      "Epoch 96/100\n",
      "804/804 [==============================] - 1s 809us/step - loss: 0.5232 - accuracy: 0.7467\n",
      "Epoch 97/100\n",
      "804/804 [==============================] - 1s 732us/step - loss: 0.5224 - accuracy: 0.7465\n",
      "Epoch 98/100\n",
      "804/804 [==============================] - 1s 816us/step - loss: 0.5224 - accuracy: 0.74710s - loss: 0.5139 - accuracy: 0.74 - ETA: 0s - loss: 0.518\n",
      "Epoch 99/100\n",
      "804/804 [==============================] - 1s 815us/step - loss: 0.5220 - accuracy: 0.7475\n",
      "Epoch 100/100\n",
      "748/804 [==========================>...] - ETA: 0s - loss: 0.5225 - accuracy: 0.7471\n",
      "Epoch 00100: saving model to checkpoints\\weights.100.hdf5\n",
      "804/804 [==============================] - 1s 769us/step - loss: 0.5224 - accuracy: 0.7469\n"
     ]
    }
   ],
   "source": [
    "# Define the checkpoint path and filenames\n",
    "os.makedirs(\"checkpoints_opt/\", exist_ok=True)\n",
    "checkpoint_path = \"checkpoints/weights.{epoch:02d}.hdf5\"\n",
    "\n",
    "# Create a callback that saves the model's weights every epoch\n",
    "cp_callback = ModelCheckpoint(\n",
    "    filepath=checkpoint_path,\n",
    "    verbose=1,\n",
    "    save_weights_only=True,\n",
    "    save_freq=\"epoch\",\n",
    "    period=5)\n",
    "\n",
    "# Train the model\n",
    "fit_model = nn.fit(X_train_scaled, y_train, epochs=100, callbacks=[cp_callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "268/268 - 0s - loss: 0.6220 - accuracy: 0.7221 - 199ms/epoch - 743us/step\n",
      "Loss: 0.6220081448554993, Accuracy: 0.7220991253852844\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model using the test data\n",
    "model_loss, model_accuracy = nn.evaluate(X_test_scaled, y_test, verbose=2)\n",
    "print(f\"Loss: {model_loss}, Accuracy: {model_accuracy}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Export our model to HDF5 file\n",
    "nn.save(\"AlphabetSoupCharity_Optimization.h5\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### - 3rd attempt : The activation function of hidden layers is changed -"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " dense_7 (Dense)             (None, 80)                3840      \n",
      "                                                                 \n",
      " dense_8 (Dense)             (None, 30)                2430      \n",
      "                                                                 \n",
      " dense_9 (Dense)             (None, 1)                 31        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 6,301\n",
      "Trainable params: 6,301\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# Define the model - deep neural net, i.e., the number of input features and hidden nodes for each layer.\n",
    "number_input_features = len(X_train_scaled[0])\n",
    "hidden_nodes_layer1 = 80\n",
    "hidden_nodes_layer2 = 30\n",
    "\n",
    "nn = tf.keras.models.Sequential()\n",
    "\n",
    "# First hidden layer\n",
    "nn.add(tf.keras.layers.Dense(units=hidden_nodes_layer1, input_dim=number_input_features, activation=\"tanh\"))\n",
    "\n",
    "# Second hidden layer\n",
    "nn.add(tf.keras.layers.Dense(units=hidden_nodes_layer2, activation=\"tanh\"))\n",
    "\n",
    "# Output layer\n",
    "nn.add(tf.keras.layers.Dense(units=1, activation=\"sigmoid\"))\n",
    "\n",
    "# Check the structure of the model\n",
    "nn.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model\n",
    "nn.compile(loss=\"binary_crossentropy\", optimizer=\"adam\", metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:`period` argument is deprecated. Please use `save_freq` to specify the frequency in number of batches seen.\n",
      "Epoch 1/100\n",
      "804/804 [==============================] - 1s 667us/step - loss: 0.5684 - accuracy: 0.7224\n",
      "Epoch 2/100\n",
      "804/804 [==============================] - 1s 698us/step - loss: 0.5532 - accuracy: 0.7308\n",
      "Epoch 3/100\n",
      "804/804 [==============================] - 1s 645us/step - loss: 0.5503 - accuracy: 0.7317\n",
      "Epoch 4/100\n",
      "804/804 [==============================] - 1s 673us/step - loss: 0.5481 - accuracy: 0.7336\n",
      "Epoch 5/100\n",
      "742/804 [==========================>...] - ETA: 0s - loss: 0.5463 - accuracy: 0.7374\n",
      "Epoch 00005: saving model to checkpoints\\weights.05.hdf5\n",
      "804/804 [==============================] - 1s 691us/step - loss: 0.5466 - accuracy: 0.7362\n",
      "Epoch 6/100\n",
      "804/804 [==============================] - 1s 650us/step - loss: 0.5451 - accuracy: 0.7361\n",
      "Epoch 7/100\n",
      "804/804 [==============================] - 1s 673us/step - loss: 0.5440 - accuracy: 0.7367\n",
      "Epoch 8/100\n",
      "804/804 [==============================] - 1s 672us/step - loss: 0.5436 - accuracy: 0.7359\n",
      "Epoch 9/100\n",
      "804/804 [==============================] - 1s 656us/step - loss: 0.5425 - accuracy: 0.7360\n",
      "Epoch 10/100\n",
      "766/804 [===========================>..] - ETA: 0s - loss: 0.5409 - accuracy: 0.7380\n",
      "Epoch 00010: saving model to checkpoints\\weights.10.hdf5\n",
      "804/804 [==============================] - 1s 680us/step - loss: 0.5415 - accuracy: 0.7376\n",
      "Epoch 11/100\n",
      "804/804 [==============================] - 1s 658us/step - loss: 0.5408 - accuracy: 0.7383\n",
      "Epoch 12/100\n",
      "804/804 [==============================] - 1s 676us/step - loss: 0.5406 - accuracy: 0.7369\n",
      "Epoch 13/100\n",
      "804/804 [==============================] - 1s 676us/step - loss: 0.5397 - accuracy: 0.7386\n",
      "Epoch 14/100\n",
      "804/804 [==============================] - 1s 657us/step - loss: 0.5389 - accuracy: 0.7389\n",
      "Epoch 15/100\n",
      "753/804 [===========================>..] - ETA: 0s - loss: 0.5375 - accuracy: 0.7397\n",
      "Epoch 00015: saving model to checkpoints\\weights.15.hdf5\n",
      "804/804 [==============================] - 1s 687us/step - loss: 0.5379 - accuracy: 0.7400\n",
      "Epoch 16/100\n",
      "804/804 [==============================] - 1s 696us/step - loss: 0.5379 - accuracy: 0.7395\n",
      "Epoch 17/100\n",
      "804/804 [==============================] - 1s 692us/step - loss: 0.5374 - accuracy: 0.7403\n",
      "Epoch 18/100\n",
      "804/804 [==============================] - 1s 657us/step - loss: 0.5367 - accuracy: 0.7400\n",
      "Epoch 19/100\n",
      "804/804 [==============================] - 1s 683us/step - loss: 0.5357 - accuracy: 0.7404\n",
      "Epoch 20/100\n",
      "786/804 [============================>.] - ETA: 0s - loss: 0.5359 - accuracy: 0.7401\n",
      "Epoch 00020: saving model to checkpoints\\weights.20.hdf5\n",
      "804/804 [==============================] - 1s 657us/step - loss: 0.5361 - accuracy: 0.7399\n",
      "Epoch 21/100\n",
      "804/804 [==============================] - 1s 665us/step - loss: 0.5352 - accuracy: 0.7410\n",
      "Epoch 22/100\n",
      "804/804 [==============================] - 1s 659us/step - loss: 0.5351 - accuracy: 0.7400\n",
      "Epoch 23/100\n",
      "804/804 [==============================] - 1s 664us/step - loss: 0.5343 - accuracy: 0.7415\n",
      "Epoch 24/100\n",
      "804/804 [==============================] - 1s 677us/step - loss: 0.5338 - accuracy: 0.7427\n",
      "Epoch 25/100\n",
      "802/804 [============================>.] - ETA: 0s - loss: 0.5333 - accuracy: 0.7422\n",
      "Epoch 00025: saving model to checkpoints\\weights.25.hdf5\n",
      "804/804 [==============================] - 1s 712us/step - loss: 0.5333 - accuracy: 0.7422\n",
      "Epoch 26/100\n",
      "804/804 [==============================] - 1s 693us/step - loss: 0.5335 - accuracy: 0.7420\n",
      "Epoch 27/100\n",
      "804/804 [==============================] - 1s 712us/step - loss: 0.5326 - accuracy: 0.7436\n",
      "Epoch 28/100\n",
      "804/804 [==============================] - 1s 697us/step - loss: 0.5325 - accuracy: 0.7421\n",
      "Epoch 29/100\n",
      "804/804 [==============================] - 1s 662us/step - loss: 0.5321 - accuracy: 0.7421\n",
      "Epoch 30/100\n",
      "767/804 [===========================>..] - ETA: 0s - loss: 0.5317 - accuracy: 0.7427\n",
      "Epoch 00030: saving model to checkpoints\\weights.30.hdf5\n",
      "804/804 [==============================] - 1s 695us/step - loss: 0.5317 - accuracy: 0.7428\n",
      "Epoch 31/100\n",
      "804/804 [==============================] - 1s 778us/step - loss: 0.5315 - accuracy: 0.7431\n",
      "Epoch 32/100\n",
      "804/804 [==============================] - 1s 660us/step - loss: 0.5312 - accuracy: 0.7432\n",
      "Epoch 33/100\n",
      "804/804 [==============================] - 1s 660us/step - loss: 0.5308 - accuracy: 0.7428\n",
      "Epoch 34/100\n",
      "804/804 [==============================] - 1s 641us/step - loss: 0.5306 - accuracy: 0.7436\n",
      "Epoch 35/100\n",
      "795/804 [============================>.] - ETA: 0s - loss: 0.5300 - accuracy: 0.7434\n",
      "Epoch 00035: saving model to checkpoints\\weights.35.hdf5\n",
      "804/804 [==============================] - 1s 652us/step - loss: 0.5304 - accuracy: 0.7435\n",
      "Epoch 36/100\n",
      "804/804 [==============================] - 1s 644us/step - loss: 0.5300 - accuracy: 0.7442\n",
      "Epoch 37/100\n",
      "804/804 [==============================] - 1s 658us/step - loss: 0.5297 - accuracy: 0.7438\n",
      "Epoch 38/100\n",
      "804/804 [==============================] - 1s 652us/step - loss: 0.5297 - accuracy: 0.7436\n",
      "Epoch 39/100\n",
      "804/804 [==============================] - 1s 673us/step - loss: 0.5297 - accuracy: 0.7447\n",
      "Epoch 40/100\n",
      "750/804 [==========================>...] - ETA: 0s - loss: 0.5289 - accuracy: 0.7451\n",
      "Epoch 00040: saving model to checkpoints\\weights.40.hdf5\n",
      "804/804 [==============================] - 1s 687us/step - loss: 0.5292 - accuracy: 0.7445\n",
      "Epoch 41/100\n",
      "804/804 [==============================] - 1s 654us/step - loss: 0.5291 - accuracy: 0.7434\n",
      "Epoch 42/100\n",
      "804/804 [==============================] - 1s 649us/step - loss: 0.5288 - accuracy: 0.7435\n",
      "Epoch 43/100\n",
      "804/804 [==============================] - 1s 684us/step - loss: 0.5283 - accuracy: 0.7458\n",
      "Epoch 44/100\n",
      "804/804 [==============================] - 1s 672us/step - loss: 0.5286 - accuracy: 0.7447\n",
      "Epoch 45/100\n",
      "788/804 [============================>.] - ETA: 0s - loss: 0.5280 - accuracy: 0.7455\n",
      "Epoch 00045: saving model to checkpoints\\weights.45.hdf5\n",
      "804/804 [==============================] - 1s 657us/step - loss: 0.5280 - accuracy: 0.7455\n",
      "Epoch 46/100\n",
      "804/804 [==============================] - 1s 645us/step - loss: 0.5281 - accuracy: 0.7446\n",
      "Epoch 47/100\n",
      "804/804 [==============================] - 1s 642us/step - loss: 0.5278 - accuracy: 0.7440\n",
      "Epoch 48/100\n",
      "804/804 [==============================] - 1s 705us/step - loss: 0.5279 - accuracy: 0.7458\n",
      "Epoch 49/100\n",
      "804/804 [==============================] - 1s 678us/step - loss: 0.5276 - accuracy: 0.7446\n",
      "Epoch 50/100\n",
      "782/804 [============================>.] - ETA: 0s - loss: 0.5278 - accuracy: 0.7455\n",
      "Epoch 00050: saving model to checkpoints\\weights.50.hdf5\n",
      "804/804 [==============================] - 1s 663us/step - loss: 0.5275 - accuracy: 0.7457\n",
      "Epoch 51/100\n",
      "804/804 [==============================] - 1s 703us/step - loss: 0.5271 - accuracy: 0.7447\n",
      "Epoch 52/100\n",
      "804/804 [==============================] - 1s 652us/step - loss: 0.5273 - accuracy: 0.7456\n",
      "Epoch 53/100\n",
      "804/804 [==============================] - 1s 675us/step - loss: 0.5268 - accuracy: 0.7460\n",
      "Epoch 54/100\n",
      "804/804 [==============================] - 1s 660us/step - loss: 0.5268 - accuracy: 0.7454\n",
      "Epoch 55/100\n",
      "803/804 [============================>.] - ETA: 0s - loss: 0.5269 - accuracy: 0.7454\n",
      "Epoch 00055: saving model to checkpoints\\weights.55.hdf5\n",
      "804/804 [==============================] - 1s 709us/step - loss: 0.5268 - accuracy: 0.7454\n",
      "Epoch 56/100\n",
      "804/804 [==============================] - 1s 688us/step - loss: 0.5263 - accuracy: 0.7455\n",
      "Epoch 57/100\n",
      "804/804 [==============================] - 1s 649us/step - loss: 0.5265 - accuracy: 0.7460\n",
      "Epoch 58/100\n",
      "804/804 [==============================] - 1s 660us/step - loss: 0.5263 - accuracy: 0.7453\n",
      "Epoch 59/100\n",
      "804/804 [==============================] - 1s 653us/step - loss: 0.5261 - accuracy: 0.7475\n",
      "Epoch 60/100\n",
      "767/804 [===========================>..] - ETA: 0s - loss: 0.5268 - accuracy: 0.7460\n",
      "Epoch 00060: saving model to checkpoints\\weights.60.hdf5\n",
      "804/804 [==============================] - 1s 678us/step - loss: 0.5259 - accuracy: 0.7469\n",
      "Epoch 61/100\n",
      "804/804 [==============================] - 1s 677us/step - loss: 0.5260 - accuracy: 0.7451\n",
      "Epoch 62/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "804/804 [==============================] - 1s 637us/step - loss: 0.5257 - accuracy: 0.7457\n",
      "Epoch 63/100\n",
      "804/804 [==============================] - 1s 658us/step - loss: 0.5256 - accuracy: 0.7457\n",
      "Epoch 64/100\n",
      "804/804 [==============================] - 1s 650us/step - loss: 0.5255 - accuracy: 0.7462\n",
      "Epoch 65/100\n",
      "772/804 [===========================>..] - ETA: 0s - loss: 0.5260 - accuracy: 0.7456\n",
      "Epoch 00065: saving model to checkpoints\\weights.65.hdf5\n",
      "804/804 [==============================] - 1s 675us/step - loss: 0.5256 - accuracy: 0.7460\n",
      "Epoch 66/100\n",
      "804/804 [==============================] - 1s 685us/step - loss: 0.5255 - accuracy: 0.7457\n",
      "Epoch 67/100\n",
      "804/804 [==============================] - 1s 671us/step - loss: 0.5252 - accuracy: 0.7455\n",
      "Epoch 68/100\n",
      "804/804 [==============================] - 1s 643us/step - loss: 0.5250 - accuracy: 0.7459\n",
      "Epoch 69/100\n",
      "804/804 [==============================] - 1s 656us/step - loss: 0.5250 - accuracy: 0.7466\n",
      "Epoch 70/100\n",
      "773/804 [===========================>..] - ETA: 0s - loss: 0.5249 - accuracy: 0.7466\n",
      "Epoch 00070: saving model to checkpoints\\weights.70.hdf5\n",
      "804/804 [==============================] - 1s 670us/step - loss: 0.5253 - accuracy: 0.7461\n",
      "Epoch 71/100\n",
      "804/804 [==============================] - 1s 681us/step - loss: 0.5247 - accuracy: 0.7462\n",
      "Epoch 72/100\n",
      "804/804 [==============================] - 1s 668us/step - loss: 0.5247 - accuracy: 0.7462\n",
      "Epoch 73/100\n",
      "804/804 [==============================] - 1s 674us/step - loss: 0.5246 - accuracy: 0.7462\n",
      "Epoch 74/100\n",
      "804/804 [==============================] - 1s 642us/step - loss: 0.5244 - accuracy: 0.7460\n",
      "Epoch 75/100\n",
      "753/804 [===========================>..] - ETA: 0s - loss: 0.5236 - accuracy: 0.7467\n",
      "Epoch 00075: saving model to checkpoints\\weights.75.hdf5\n",
      "804/804 [==============================] - 1s 690us/step - loss: 0.5246 - accuracy: 0.7463\n",
      "Epoch 76/100\n",
      "804/804 [==============================] - 1s 672us/step - loss: 0.5243 - accuracy: 0.7450\n",
      "Epoch 77/100\n",
      "804/804 [==============================] - 1s 649us/step - loss: 0.5245 - accuracy: 0.7470\n",
      "Epoch 78/100\n",
      "804/804 [==============================] - 1s 673us/step - loss: 0.5243 - accuracy: 0.7466\n",
      "Epoch 79/100\n",
      "804/804 [==============================] - 1s 656us/step - loss: 0.5242 - accuracy: 0.7462\n",
      "Epoch 80/100\n",
      "790/804 [============================>.] - ETA: 0s - loss: 0.5249 - accuracy: 0.7453\n",
      "Epoch 00080: saving model to checkpoints\\weights.80.hdf5\n",
      "804/804 [==============================] - 1s 655us/step - loss: 0.5241 - accuracy: 0.7460\n",
      "Epoch 81/100\n",
      "804/804 [==============================] - 1s 656us/step - loss: 0.5239 - accuracy: 0.7466\n",
      "Epoch 82/100\n",
      "804/804 [==============================] - 1s 642us/step - loss: 0.5241 - accuracy: 0.7463\n",
      "Epoch 83/100\n",
      "804/804 [==============================] - 1s 656us/step - loss: 0.5237 - accuracy: 0.7453\n",
      "Epoch 84/100\n",
      "804/804 [==============================] - 1s 641us/step - loss: 0.5240 - accuracy: 0.7465\n",
      "Epoch 85/100\n",
      "789/804 [============================>.] - ETA: 0s - loss: 0.5235 - accuracy: 0.7468\n",
      "Epoch 00085: saving model to checkpoints\\weights.85.hdf5\n",
      "804/804 [==============================] - 1s 658us/step - loss: 0.5236 - accuracy: 0.7467\n",
      "Epoch 86/100\n",
      "804/804 [==============================] - 1s 652us/step - loss: 0.5236 - accuracy: 0.7460\n",
      "Epoch 87/100\n",
      "804/804 [==============================] - 1s 654us/step - loss: 0.5235 - accuracy: 0.7470\n",
      "Epoch 88/100\n",
      "804/804 [==============================] - 1s 647us/step - loss: 0.5235 - accuracy: 0.7463\n",
      "Epoch 89/100\n",
      "804/804 [==============================] - 1s 656us/step - loss: 0.5235 - accuracy: 0.7471\n",
      "Epoch 90/100\n",
      "790/804 [============================>.] - ETA: 0s - loss: 0.5230 - accuracy: 0.7470\n",
      "Epoch 00090: saving model to checkpoints\\weights.90.hdf5\n",
      "804/804 [==============================] - 1s 660us/step - loss: 0.5235 - accuracy: 0.7467\n",
      "Epoch 91/100\n",
      "804/804 [==============================] - 1s 644us/step - loss: 0.5234 - accuracy: 0.7463\n",
      "Epoch 92/100\n",
      "804/804 [==============================] - 1s 655us/step - loss: 0.5233 - accuracy: 0.7456\n",
      "Epoch 93/100\n",
      "804/804 [==============================] - 1s 639us/step - loss: 0.5229 - accuracy: 0.7478\n",
      "Epoch 94/100\n",
      "804/804 [==============================] - 1s 668us/step - loss: 0.5230 - accuracy: 0.7457\n",
      "Epoch 95/100\n",
      "775/804 [===========================>..] - ETA: 0s - loss: 0.5226 - accuracy: 0.7471\n",
      "Epoch 00095: saving model to checkpoints\\weights.95.hdf5\n",
      "804/804 [==============================] - 1s 667us/step - loss: 0.5230 - accuracy: 0.7463\n",
      "Epoch 96/100\n",
      "804/804 [==============================] - 1s 674us/step - loss: 0.5231 - accuracy: 0.7468\n",
      "Epoch 97/100\n",
      "804/804 [==============================] - 1s 675us/step - loss: 0.5228 - accuracy: 0.7467\n",
      "Epoch 98/100\n",
      "804/804 [==============================] - 1s 671us/step - loss: 0.5229 - accuracy: 0.7467\n",
      "Epoch 99/100\n",
      "804/804 [==============================] - 1s 643us/step - loss: 0.5227 - accuracy: 0.7468\n",
      "Epoch 100/100\n",
      "752/804 [===========================>..] - ETA: 0s - loss: 0.5229 - accuracy: 0.7466\n",
      "Epoch 00100: saving model to checkpoints\\weights.100.hdf5\n",
      "804/804 [==============================] - 1s 683us/step - loss: 0.5227 - accuracy: 0.7470\n"
     ]
    }
   ],
   "source": [
    "# Define the checkpoint path and filenames\n",
    "os.makedirs(\"checkpoints_opt/\", exist_ok=True)\n",
    "checkpoint_path = \"checkpoints/weights.{epoch:02d}.hdf5\"\n",
    "\n",
    "# Create a callback that saves the model's weights every epoch\n",
    "cp_callback = ModelCheckpoint(\n",
    "    filepath=checkpoint_path,\n",
    "    verbose=1,\n",
    "    save_weights_only=True,\n",
    "    save_freq=\"epoch\",\n",
    "    period=5)\n",
    "\n",
    "# Train the model\n",
    "fit_model = nn.fit(X_train_scaled, y_train, epochs=100, callbacks=[cp_callback])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "268/268 - 0s - loss: 0.5695 - accuracy: 0.7214 - 182ms/epoch - 678us/step\n",
      "Loss: 0.5694616436958313, Accuracy: 0.7213994264602661\n"
     ]
    }
   ],
   "source": [
    "# Evaluate the model using the test data\n",
    "model_loss, model_accuracy = nn.evaluate(X_test_scaled, y_test, verbose=2)\n",
    "print(f\"Loss: {model_loss}, Accuracy: {model_accuracy}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Export our model to HDF5 file\n",
    "nn.save(\"AlphabetSoupCharity_Optimization.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "mlenv",
   "language": "python",
   "name": "mlenv"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}