3. modeling_performa_karyawan_resign_cluster_DL.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6843e881",
   "metadata": {},
   "source": [
    "# Employee Performance With Resign Prediction Using Deep Learning (3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "9c9de905",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "065acde8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# library dibawah sebenarnya udah ada di pyforest\n",
    "from sklearn.model_selection       import train_test_split\n",
    "from sklearn.feature_selection     import SelectKBest, f_classif\n",
    "from sklearn.preprocessing         import MinMaxScaler, StandardScaler\n",
    "from sklearn.preprocessing         import LabelEncoder, OrdinalEncoder, OneHotEncoder\n",
    "from sklearn.pipeline              import make_pipeline\n",
    "from sklearn.compose               import ColumnTransformer\n",
    "from sklearn.metrics               import (f1_score,\n",
    "                                           precision_score,\n",
    "                                           recall_score,\n",
    "                                           log_loss,\n",
    "                                           accuracy_score,\n",
    "                                           classification_report,\n",
    "                                           ConfusionMatrixDisplay,\n",
    "                                           confusion_matrix)\n",
    "\n",
    "\n",
    "from pyspark.sql                   import SparkSession\n",
    "from sklearn.ensemble              import StackingClassifier\n",
    "from sklearn.pipeline              import Pipeline\n",
    "from sklearn.preprocessing         import StandardScaler\n",
    "from sklearn_pandas                import DataFrameMapper\n",
    "from sklearn.ensemble              import AdaBoostClassifier\n",
    "from sklearn.tree                  import DecisionTreeClassifier\n",
    "from sklearn.linear_model          import LogisticRegression\n",
    "from matplotlib.colors             import ListedColormap\n",
    "from sklearn.linear_model          import SGDClassifier\n",
    "from sklearn.ensemble              import RandomForestClassifier\n",
    "from sklearn.linear_model          import SGDClassifier\n",
    "from sklearn.svm                   import SVC\n",
    "from sklearn.naive_bayes           import GaussianNB\n",
    "from sklearn.neighbors             import KNeighborsClassifier\n",
    "from sklearn.neighbors             import NeighborhoodComponentsAnalysis\n",
    "from keras.callbacks               import EarlyStopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f9212281",
   "metadata": {},
   "outputs": [],
   "source": [
    "# env\n",
    "os.environ['HADOOP_HOME'] = r\"C:\\hadoop\"\n",
    "os.environ['HADOOP_CONF_DIR'] = r\"C:\\hadoop\\etc\\hadoop\"\n",
    "os.environ['JAVA_HOME'] = r\"C:\\Java\"  # path jdk\n",
    "\n",
    "# sesi spark\n",
    "spark = SparkSession.builder \\\n",
    "    .appName(\"Hadoop Clustering Performa Karyawan\") \\\n",
    "    .config(\"spark.hadoop.fs.defaultFS\", \"hdfs://localhost:9000\") \\\n",
    "    .config(\"spark.yarn.stagingDir\", \"/user/ASUS/.sparkStaging\") \\\n",
    "    .getOrCreate()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a6cfcc07",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# data hdfs\n",
    "hdfs_path = \"/user/ASUS/hadoop_clustering_employee_turnover/cluster.csv\"\n",
    "data_hadoop = spark.read.csv(hdfs_path, header=True, inferSchema=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "1723da30",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df = pd.DataFrame(data_hadoop)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "bbc6b508",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 14999 entries, 0 to 14998\n",
      "Data columns (total 20 columns):\n",
      " #   Column                      Non-Null Count  Dtype  \n",
      "---  ------                      --------------  -----  \n",
      " 0   tingkat_kepuasan            14999 non-null  float64\n",
      " 1   evaluasi_terakhir           14999 non-null  float64\n",
      " 2   jumlah_proyek               14999 non-null  int64  \n",
      " 3   rata-rata_jam_bulan         14999 non-null  int64  \n",
      " 4   waktu_spent_for_perusahaan  14999 non-null  int64  \n",
      " 5   Kecelakaan_kerja            14999 non-null  int64  \n",
      " 6   resign                      14999 non-null  int64  \n",
      " 7   promosi_5tahun_terakhir     14999 non-null  int64  \n",
      " 8   gaji                        14999 non-null  int64  \n",
      " 9   IT                          14999 non-null  int64  \n",
      " 10  RandD                       14999 non-null  int64  \n",
      " 11  accounting                  14999 non-null  int64  \n",
      " 12  hr                          14999 non-null  int64  \n",
      " 13  management                  14999 non-null  int64  \n",
      " 14  marketing                   14999 non-null  int64  \n",
      " 15  product_mng                 14999 non-null  int64  \n",
      " 16  sales                       14999 non-null  int64  \n",
      " 17  support                     14999 non-null  int64  \n",
      " 18  technical                   14999 non-null  int64  \n",
      " 19  cluster                     14999 non-null  int64  \n",
      "dtypes: float64(2), int64(18)\n",
      "memory usage: 2.3 MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "23ab47dc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = df.drop('resign', axis=1)\n",
    "y = pd.DataFrame(df['resign'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ab633198",
   "metadata": {},
   "outputs": [],
   "source": [
    "hapus = ['accounting',\n",
    "        'product_mng',\n",
    "        'IT',\n",
    "        'support',\n",
    "        'sales',\n",
    "        'evaluasi_terakhir',\n",
    "        'marketing']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "0bb2b497",
   "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>tingkat_kepuasan</th>\n",
       "      <th>jumlah_proyek</th>\n",
       "      <th>rata-rata_jam_bulan</th>\n",
       "      <th>waktu_spent_for_perusahaan</th>\n",
       "      <th>Kecelakaan_kerja</th>\n",
       "      <th>promosi_5tahun_terakhir</th>\n",
       "      <th>gaji</th>\n",
       "      <th>RandD</th>\n",
       "      <th>hr</th>\n",
       "      <th>management</th>\n",
       "      <th>technical</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.38</td>\n",
       "      <td>2</td>\n",
       "      <td>157</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.80</td>\n",
       "      <td>5</td>\n",
       "      <td>262</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.11</td>\n",
       "      <td>7</td>\n",
       "      <td>272</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.72</td>\n",
       "      <td>5</td>\n",
       "      <td>223</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.37</td>\n",
       "      <td>2</td>\n",
       "      <td>159</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14994</th>\n",
       "      <td>0.40</td>\n",
       "      <td>2</td>\n",
       "      <td>151</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14995</th>\n",
       "      <td>0.37</td>\n",
       "      <td>2</td>\n",
       "      <td>160</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14996</th>\n",
       "      <td>0.37</td>\n",
       "      <td>2</td>\n",
       "      <td>143</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14997</th>\n",
       "      <td>0.11</td>\n",
       "      <td>6</td>\n",
       "      <td>280</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14998</th>\n",
       "      <td>0.37</td>\n",
       "      <td>2</td>\n",
       "      <td>158</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>14999 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       tingkat_kepuasan  jumlah_proyek  rata-rata_jam_bulan  ...  management  technical  cluster\n",
       "0                  0.38              2                  157  ...           0          0        2\n",
       "1                  0.80              5                  262  ...           0          0        3\n",
       "2                  0.11              7                  272  ...           0          0        3\n",
       "3                  0.72              5                  223  ...           0          0        0\n",
       "4                  0.37              2                  159  ...           0          0        2\n",
       "...                 ...            ...                  ...  ...         ...        ...      ...\n",
       "14994              0.40              2                  151  ...           0          0        2\n",
       "14995              0.37              2                  160  ...           0          0        2\n",
       "14996              0.37              2                  143  ...           0          0        4\n",
       "14997              0.11              6                  280  ...           0          0        3\n",
       "14998              0.37              2                  158  ...           0          0        2\n",
       "\n",
       "[14999 rows x 12 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = X.drop(hapus, axis=1).sort_index()\n",
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b780bbc4",
   "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>resign</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14994</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14995</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14996</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14997</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14998</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>14999 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       resign\n",
       "0           1\n",
       "1           1\n",
       "2           1\n",
       "3           1\n",
       "4           1\n",
       "...       ...\n",
       "14994       1\n",
       "14995       1\n",
       "14996       1\n",
       "14997       1\n",
       "14998       1\n",
       "\n",
       "[14999 rows x 1 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "cc951a59",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "resign    int64\n",
       "dtype: object"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "55874204",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X,\n",
    "                                                    y,\n",
    "                                                    test_size=0.25,\n",
    "                                                    random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "8d741b2f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(11249, 12)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "26a546c0",
   "metadata": {},
   "source": [
    "### Neural Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "655c4683",
   "metadata": {},
   "outputs": [],
   "source": [
    "# y_true itu y_test\n",
    "def model_report(y_true, y_pred):\n",
    "\n",
    "    acc                     = accuracy_score(y_true, y_pred)\n",
    "    pre                     = precision_score(y_true, y_pred)\n",
    "    rec                     = recall_score(y_true, y_pred)\n",
    "    f1                      = f1_score(y_true, y_pred)\n",
    "    loss                    = log_loss(y_true, y_pred)\n",
    "\n",
    "    print(\"accuracy:   {:3f}\\nprecision:  {:3f}\\nrecall:     {:3f}\\nf1-score:   {:3f}\\nloss:       {:3f}\".\n",
    "          format(acc, pre, rec, f1, loss))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "6c3aedd3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Fungsi Evaluasi Suatu Model dengan Confusion Matrix\n",
    "def plot_confusion_matriks(y_true, y_pred):\n",
    "    cm    = confusion_matrix(y_true, y_pred)\n",
    "    tampil = ConfusionMatrixDisplay(confusion_matrix=cm)\n",
    "\n",
    "    tampil.plot(cmap='Greys')\n",
    "\n",
    "    plt.title('Confusion Matrix')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "8ddd5ece",
   "metadata": {},
   "outputs": [],
   "source": [
    "def pred_report_nn(y_true, y_pred):\n",
    "\n",
    "    # Konversi prediksi yang kontinu menjadi nilai biner (0/1)\n",
    "    y_pred_bin = np.round(y_pred).astype(int)\n",
    "\n",
    "    # Hitung metrik evaluasi klasifikasi\n",
    "    acc_ann    = accuracy_score(y_true, y_pred_bin)\n",
    "    pre_ann    = precision_score(y_true, y_pred_bin)\n",
    "    rec_ann    = recall_score(y_true, y_pred_bin)\n",
    "    f1_ann     = f1_score(y_true, y_pred_bin)\n",
    "    loss_ann   = log_loss(y_true, y_pred)\n",
    "\n",
    "    print(\"Accuracy:  {:3f}\\nPrecision:  {:3f}\\nRecall:  {:3f}\\nF1 Score:  {:3f}\\nLoss:  {:3f}\".\n",
    "          format(acc_ann, pre_ann, rec_ann, f1_ann, loss_ann))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1abf2df2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(11249, 12)\n",
      "(3750, 12)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape)\n",
    "print(X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "9d198b34",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "early_stopping = keras.callbacks.EarlyStopping(monitor='val_loss',\n",
    "                                               patience=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "08f2b079",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ann_model = keras.Sequential()\n",
    "\n",
    "ann_model.add(keras.layers.Dense(units=24, input_shape=(12,)))\n",
    "ann_model.add(keras.layers.LeakyReLU(alpha=0.01))\n",
    "ann_model.add(keras.layers.Dropout(0.1)) #untuk antisipasi overfit\n",
    "\n",
    "ann_model.add(keras.layers.Dense(units=12))\n",
    "ann_model.add(keras.layers.LeakyReLU(alpha=0.01))\n",
    "ann_model.add(keras.layers.Dropout(0.2))\n",
    "\n",
    "ann_model.add(keras.layers.Dense(units=6))\n",
    "ann_model.add(keras.layers.LeakyReLU(alpha=0.01))\n",
    "ann_model.add(keras.layers.Dropout(0.15))\n",
    "\n",
    "ann_model.add(keras.layers.Dense(units=1, activation=\"sigmoid\")) #output\n",
    "\n",
    "ann_model.compile(optimizer=keras.optimizers.AdamW(0.01), #learning rate\n",
    "                 loss=\"mse\",\n",
    "                 metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "078aecb4",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "282/282 - 5s - 17ms/step - accuracy: 0.7563 - loss: 0.2356 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 2/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7609 - loss: 0.2322 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 3/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7609 - loss: 0.2306 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 4/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2297 - val_accuracy: 0.7671 - val_loss: 0.1790\n",
      "Epoch 5/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.1829 - val_accuracy: 0.7671 - val_loss: 0.1786\n",
      "Epoch 6/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7611 - loss: 0.1819 - val_accuracy: 0.7671 - val_loss: 0.1786\n",
      "Epoch 7/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.1819 - val_accuracy: 0.7671 - val_loss: 0.1784\n",
      "Epoch 8/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7611 - loss: 0.1817 - val_accuracy: 0.7671 - val_loss: 0.1783\n",
      "Epoch 9/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.1814 - val_accuracy: 0.7671 - val_loss: 0.1777\n",
      "Epoch 10/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7610 - loss: 0.1803 - val_accuracy: 0.7671 - val_loss: 0.1728\n",
      "Epoch 11/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7613 - loss: 0.2265 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 12/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7610 - loss: 0.2324 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 13/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7612 - loss: 0.2310 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 14/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2334 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 15/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7605 - loss: 0.2386 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 16/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2383 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 17/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7611 - loss: 0.2385 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 18/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7610 - loss: 0.2373 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 19/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7634 - loss: 0.1851 - val_accuracy: 0.7791 - val_loss: 0.1616\n",
      "Epoch 20/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7729 - loss: 0.1622 - val_accuracy: 0.8111 - val_loss: 0.1427\n",
      "Epoch 21/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7763 - loss: 0.1562 - val_accuracy: 0.7871 - val_loss: 0.1406\n",
      "Epoch 22/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7920 - loss: 0.1487 - val_accuracy: 0.7653 - val_loss: 0.1501\n",
      "Epoch 23/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7969 - loss: 0.1449 - val_accuracy: 0.8311 - val_loss: 0.1214\n",
      "Epoch 24/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.8232 - loss: 0.1286 - val_accuracy: 0.8516 - val_loss: 0.1090\n",
      "Epoch 25/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7868 - loss: 0.1944 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 26/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2390 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 27/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7611 - loss: 0.2389 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 28/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2389 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 29/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2388 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 30/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2389 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 31/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2388 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 32/50\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7611 - loss: 0.2389 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 33/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2389 - val_accuracy: 0.7671 - val_loss: 0.2329\n",
      "Epoch 34/50\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.2388 - val_accuracy: 0.7671 - val_loss: 0.2329\n"
     ]
    }
   ],
   "source": [
    "ann_latih = ann_model.fit(X_train, y_train,\n",
    "                         epochs=50,\n",
    "                         validation_split=0.2,\n",
    "                         callbacks=[early_stopping],\n",
    "                         verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "28568b83",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ann_model3 = keras.Sequential()\n",
    "\n",
    "ann_model3.add(keras.layers.Dense(units=64, activation=\"relu\", input_shape=(12,)))\n",
    "ann_model3.add(keras.layers.Dropout(0.3))\n",
    "\n",
    "ann_model3.add(keras.layers.Dense(units=32, activation=\"relu\"))\n",
    "ann_model3.add(keras.layers.Dropout(0.3))\n",
    "\n",
    "ann_model3.add(keras.layers.Dense(units=16, activation=\"relu\"))\n",
    "ann_model3.add(keras.layers.Dropout(0.2))\n",
    "\n",
    "ann_model3.add(keras.layers.Dense(units=1, activation=\"sigmoid\"))\n",
    "\n",
    "ann_model3.compile(optimizer=keras.optimizers.RMSprop(learning_rate=0.001), \n",
    "                   loss=\"binary_crossentropy\",\n",
    "                   metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "c24a4b14",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "282/282 - 4s - 14ms/step - accuracy: 0.6430 - loss: 1.8505 - val_accuracy: 0.7671 - val_loss: 0.6345\n",
      "Epoch 2/100\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7479 - loss: 0.6130 - val_accuracy: 0.7671 - val_loss: 0.5800\n",
      "Epoch 3/100\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7582 - loss: 0.5801 - val_accuracy: 0.7671 - val_loss: 0.5594\n",
      "Epoch 4/100\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7601 - loss: 0.5674 - val_accuracy: 0.7671 - val_loss: 0.5485\n",
      "Epoch 5/100\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7608 - loss: 0.5630 - val_accuracy: 0.7671 - val_loss: 0.5461\n",
      "Epoch 6/100\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7608 - loss: 0.5563 - val_accuracy: 0.7671 - val_loss: 0.5463\n",
      "Epoch 7/100\n",
      "282/282 - 1s - 5ms/step - accuracy: 0.7610 - loss: 0.5558 - val_accuracy: 0.7671 - val_loss: 0.5467\n",
      "Epoch 8/100\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7612 - loss: 0.5445 - val_accuracy: 0.7671 - val_loss: 0.5248\n",
      "Epoch 9/100\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.5312 - val_accuracy: 0.7671 - val_loss: 0.5030\n",
      "Epoch 10/100\n",
      "282/282 - 1s - 4ms/step - accuracy: 0.7611 - loss: 0.4909 - val_accuracy: 0.7671 - val_loss: 0.4262\n"
     ]
    }
   ],
   "source": [
    "ann_latih3 = ann_model3.fit(X_train, y_train,\n",
    "                            epochs=100,\n",
    "                            validation_split=0.2,\n",
    "                            callbacks=[early_stopping],\n",
    "                            verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "585ec5b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.callbacks import LearningRateScheduler\n",
    "\n",
    "def learning_rate_scheduler(epoch, lr):\n",
    "    \n",
    "    if epoch % 10 == 0 and epoch != 0:\n",
    "        lr = lr * 0.1\n",
    "        \n",
    "    return lr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "0938efc8",
   "metadata": {},
   "outputs": [],
   "source": [
    "lr_schedul = LearningRateScheduler(learning_rate_scheduler)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "c56a2b7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ann_model4 = keras.Sequential()\n",
    "\n",
    "ann_model4.add(keras.layers.Dense(units=12, activation=\"relu\", input_shape=(12,)))\n",
    "ann_model4.add(keras.layers.Dropout(0.3))\n",
    "\n",
    "ann_model4.add(keras.layers.Dense(units=6, activation=\"relu\"))\n",
    "ann_model4.add(keras.layers.Dropout(0.2))\n",
    "\n",
    "ann_model4.add(keras.layers.Dense(units=1, activation=\"sigmoid\"))\n",
    "\n",
    "ann_model4.compile(optimizer=keras.optimizers.RMSprop(learning_rate=0.01), \n",
    "                   loss=\"binary_crossentropy\",\n",
    "                   metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "e367b922",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "71/71 - 3s - 38ms/step - accuracy: 0.7239 - loss: 1.6885 - val_accuracy: 0.7671 - val_loss: 0.5690 - learning_rate: 0.0100\n",
      "Epoch 2/100\n",
      "71/71 - 0s - 5ms/step - accuracy: 0.7589 - loss: 0.5599 - val_accuracy: 0.7671 - val_loss: 0.5438 - learning_rate: 0.0100\n",
      "Epoch 3/100\n",
      "71/71 - 0s - 5ms/step - accuracy: 0.7611 - loss: 0.5499 - val_accuracy: 0.7671 - val_loss: 0.5427 - learning_rate: 0.0100\n",
      "Epoch 4/100\n",
      "71/71 - 0s - 5ms/step - accuracy: 0.7611 - loss: 0.5504 - val_accuracy: 0.7671 - val_loss: 0.5431 - learning_rate: 0.0100\n",
      "Epoch 5/100\n",
      "71/71 - 0s - 5ms/step - accuracy: 0.7611 - loss: 0.5502 - val_accuracy: 0.7671 - val_loss: 0.5427 - learning_rate: 0.0100\n",
      "Epoch 6/100\n",
      "71/71 - 0s - 5ms/step - accuracy: 0.7611 - loss: 0.5500 - val_accuracy: 0.7671 - val_loss: 0.5429 - learning_rate: 0.0100\n",
      "Epoch 7/100\n",
      "71/71 - 0s - 5ms/step - accuracy: 0.7611 - loss: 0.5495 - val_accuracy: 0.7671 - val_loss: 0.5435 - learning_rate: 0.0100\n",
      "Epoch 8/100\n",
      "71/71 - 0s - 5ms/step - accuracy: 0.7611 - loss: 0.5502 - val_accuracy: 0.7671 - val_loss: 0.5429 - learning_rate: 0.0100\n",
      "Epoch 9/100\n",
      "71/71 - 0s - 5ms/step - accuracy: 0.7611 - loss: 0.5501 - val_accuracy: 0.7671 - val_loss: 0.5432 - learning_rate: 0.0100\n",
      "Epoch 10/100\n",
      "71/71 - 1s - 10ms/step - accuracy: 0.7611 - loss: 0.5500 - val_accuracy: 0.7671 - val_loss: 0.5429 - learning_rate: 0.0100\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.history.History at 0x27254dffa10>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ann_model4.fit(X_train,\n",
    "               y_train,\n",
    "               epochs=100,\n",
    "               batch_size=128,\n",
    "               validation_split=0.2,\n",
    "               callbacks=[lr_schedul, early_stopping],\n",
    "               verbose=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "e77e66ed",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ann_model7 = keras.Sequential()\n",
    "\n",
    "ann_model7.add(keras.layers.Dense(units=24, activation=\"softplus\", input_shape=(12,)))\n",
    "ann_model7.add(keras.layers.Dropout(0.1)) #untuk antisipasi overfit\n",
    "\n",
    "ann_model7.add(keras.layers.Dense(units=12, activation=\"softplus\", input_shape=(12,)))\n",
    "ann_model7.add(keras.layers.Dropout(0.1)) #untuk antisipasi overfit\n",
    "\n",
    "ann_model7.add(keras.layers.Dense(units=6, activation=\"softplus\", input_shape=(12,)))\n",
    "ann_model7.add(keras.layers.Dropout(0.1)) #untuk antisipasi overfit\n",
    "\n",
    "ann_model7.add(keras.layers.Dense(units=1, activation=\"sigmoid\")) #output\n",
    "\n",
    "ann_model7.compile(optimizer=tf.keras.optimizers.Adadelta(), \n",
    "                   loss=\"binary_crossentropy\",\n",
    "                   metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "288b255d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "282/282 - 2s - 7ms/step - accuracy: 0.4246 - loss: 10.8377 - val_accuracy: 0.2329 - val_loss: 9.5421\n",
      "Epoch 2/50\n",
      "282/282 - 1s - 2ms/step - accuracy: 0.4289 - loss: 10.4974 - val_accuracy: 0.2329 - val_loss: 8.7141\n",
      "Epoch 3/50\n",
      "282/282 - 1s - 2ms/step - accuracy: 0.4370 - loss: 10.0323 - val_accuracy: 0.2329 - val_loss: 7.8557\n",
      "Epoch 4/50\n",
      "282/282 - 1s - 2ms/step - accuracy: 0.4492 - loss: 9.3620 - val_accuracy: 0.2329 - val_loss: 6.9378\n",
      "Epoch 5/50\n",
      "282/282 - 1s - 2ms/step - accuracy: 0.4423 - loss: 9.1087 - val_accuracy: 0.2329 - val_loss: 5.9587\n",
      "Epoch 6/50\n",
      "282/282 - 1s - 2ms/step - accuracy: 0.4584 - loss: 8.4803 - val_accuracy: 0.2329 - val_loss: 5.0147\n",
      "Epoch 7/50\n",
      "282/282 - 1s - 2ms/step - accuracy: 0.4586 - loss: 8.2046 - val_accuracy: 0.2329 - val_loss: 4.0780\n",
      "Epoch 8/50\n",
      "282/282 - 1s - 2ms/step - accuracy: 0.4633 - loss: 7.9146 - val_accuracy: 0.2329 - val_loss: 3.1636\n",
      "Epoch 9/50\n",
      "282/282 - 1s - 3ms/step - accuracy: 0.4719 - loss: 7.4315 - val_accuracy: 0.2329 - val_loss: 2.3158\n",
      "Epoch 10/50\n",
      "282/282 - 1s - 2ms/step - accuracy: 0.4812 - loss: 7.0447 - val_accuracy: 0.2329 - val_loss: 1.5948\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.src.callbacks.history.History at 0x2725ab01450>"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ann_model7.fit(X_train,\n",
    "               y_train,\n",
    "               epochs=50,\n",
    "               validation_split=0.2,\n",
    "               callbacks=[early_stopping],\n",
    "               verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b3956a92",
   "metadata": {},
   "source": [
    "### First Model is ann_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "00222a2a",
   "metadata": {},
   "outputs": [],
   "source": [
    "def viz_hasil_churn(ann_latih):\n",
    "\n",
    "    fig, (ax1, ax2) = plt.subplots(1, 2)\n",
    "\n",
    "    ax1.plot(ann_latih.history[\"accuracy\"],     label= \"Model accuracy\")\n",
    "    ax1.plot(ann_latih.history[\"val_accuracy\"], label= \"Valid accuracy\")\n",
    "    ax1.set_title(\"Model accuracy\")\n",
    "    ax1.set_ylabel(\"accuracy\")\n",
    "    ax1.set_xlabel(\"epoch\")\n",
    "    \n",
    "    ax2.plot(ann_latih.history[\"loss\"],     label= \"Model loss\")\n",
    "    ax2.plot(ann_latih.history[\"val_loss\"], label= \"Valid loss\")\n",
    "    ax2.set_title(\"Model loss\", )\n",
    "    ax2.set_ylabel(\"loss\")\n",
    "    ax2.set_xlabel(\"epoch\")\n",
    "\n",
    "    ax1.legend()\n",
    "    ax2.legend()\n",
    "    \n",
    "    ax1.spines[['right', 'top']] .set_visible(False)\n",
    "    ax2.spines[['right', 'top']] .set_visible(False)\n",
    "    \n",
    "\n",
    "    fig.suptitle(\"Model loss & accuracy\")\n",
    "    fig.set_size_inches(12, 6)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "efac6c66",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport matplotlib.pyplot as plt\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import pandas as pd\\nimport keras\\nimport matplotlib.pyplot as plt\\nimport tensorflow as tf'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "viz_hasil_churn(ann_latih)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "aefb2e67",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Hasil Evaluasi Model terhadap Prediksi data test menggunakan NN:\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "        if (window._pyforest_update_imports_cell) { window._pyforest_update_imports_cell('import tensorflow as tf\\nimport matplotlib.pyplot as plt\\nimport numpy as np\\nimport keras\\nimport pandas as pd'); }\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy:  0.760800\n",
      "Precision:  0.000000\n",
      "Recall:  0.000000\n",
      "F1 Score:  0.000000\n",
      "Loss:  3.813419\n"
     ]
    }
   ],
   "source": [
    "print(\"Hasil Evaluasi Model terhadap Prediksi data test menggunakan NN:\")\n",
    "pred_report_nn(y_test, ann_model.predict(X_test, verbose=0))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1388cf32",
   "metadata": {},
   "source": [
    "### Seems like this model is not good and doesnt fit to use when deploy to an app. so yeah i think Light GBM is still the winner !!"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}