Predicting House Price using Neural Network.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "65020dda",
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.datasets import boston_housing\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from keras import models\n",
    "from keras import layers\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from sklearn.model_selection import KFold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "0dcb4652",
   "metadata": {},
   "outputs": [],
   "source": [
    "(train_data, train_targets), (test_data, test_targets) = boston_housing.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "d138d438",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.80846e+01, 0.00000e+00, 1.81000e+01, ..., 2.02000e+01,\n",
       "        2.72500e+01, 2.90500e+01],\n",
       "       [1.23290e-01, 0.00000e+00, 1.00100e+01, ..., 1.78000e+01,\n",
       "        3.94950e+02, 1.62100e+01],\n",
       "       [5.49700e-02, 0.00000e+00, 5.19000e+00, ..., 2.02000e+01,\n",
       "        3.96900e+02, 9.74000e+00],\n",
       "       ...,\n",
       "       [1.83377e+00, 0.00000e+00, 1.95800e+01, ..., 1.47000e+01,\n",
       "        3.89610e+02, 1.92000e+00],\n",
       "       [3.58090e-01, 0.00000e+00, 6.20000e+00, ..., 1.74000e+01,\n",
       "        3.91700e+02, 9.71000e+00],\n",
       "       [2.92400e+00, 0.00000e+00, 1.95800e+01, ..., 1.47000e+01,\n",
       "        2.40160e+02, 9.81000e+00]])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "c207c80d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((404, 13), (102, 13))"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.shape, test_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "c3c70969",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([15.2, 42.3, 50. , 21.1, 17.7, 18.5, 11.3, 15.6, 15.6, 14.4])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_targets[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "9aa4a1a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_np_array = np.array(train_targets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "f4126cc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_target = pd.DataFrame(my_np_array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "ed199b57",
   "metadata": {
    "scrolled": true
   },
   "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>404.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>22.395050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>9.210442</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>16.675000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>20.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>24.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>50.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                0\n",
       "count  404.000000\n",
       "mean    22.395050\n",
       "std      9.210442\n",
       "min      5.000000\n",
       "25%     16.675000\n",
       "50%     20.750000\n",
       "75%     24.800000\n",
       "max     50.000000"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_target.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c097cf1",
   "metadata": {},
   "source": [
    "## Manually Normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "aaca4650",
   "metadata": {},
   "outputs": [],
   "source": [
    "mean = train_data.mean(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "b5069e7f",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([3.74511057e+00, 1.14801980e+01, 1.11044307e+01, 6.18811881e-02,\n",
       "       5.57355941e-01, 6.26708168e+00, 6.90106436e+01, 3.74027079e+00,\n",
       "       9.44059406e+00, 4.05898515e+02, 1.84759901e+01, 3.54783168e+02,\n",
       "       1.27408168e+01])"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "197651ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data-=mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "99101e67",
   "metadata": {},
   "outputs": [],
   "source": [
    "std = train_data.std(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "8d1c2882",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data/=std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "009f810f",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data-=mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "f832e3d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data/=std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "51d3fd32",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.55369355, -0.48361547,  1.0283258 , ...,  0.78447637,\n",
       "        -3.48459553,  2.25092074],\n",
       "       [-0.39242675, -0.48361547, -0.16087773, ..., -0.30759583,\n",
       "         0.42733126,  0.47880119],\n",
       "       [-0.39982927, -0.48361547, -0.86940196, ...,  0.78447637,\n",
       "         0.44807713, -0.41415936],\n",
       "       ...,\n",
       "       [-0.20709507, -0.48361547,  1.24588095, ..., -1.71818909,\n",
       "         0.37051949, -1.49344089],\n",
       "       [-0.36698601, -0.48361547, -0.72093526, ..., -0.48960787,\n",
       "         0.39275481, -0.41829982],\n",
       "       [-0.0889679 , -0.48361547,  1.24588095, ..., -1.71818909,\n",
       "        -1.21946544, -0.40449827]])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5dc6948",
   "metadata": {},
   "source": [
    "## Normalization using MinMaxScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "548169dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "(train_data, train_targets), (test_data, test_targets) = boston_housing.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "55deadde",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a MinMaxScaler object\n",
    "scaler = MinMaxScaler()\n",
    "\n",
    "# Fit the scaler on the training data\n",
    "scaler.fit(train_data)\n",
    "\n",
    "# Normalize the training data\n",
    "train_data_normalized = scaler.transform(train_data)\n",
    "\n",
    "# Normalize the test data using the same scaler\n",
    "test_data_normalized = scaler.transform(test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "4e6bb333",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.37816304e-02 0.00000000e+00 2.81524927e-01 ... 8.93617021e-01\n",
      "  1.00000000e+00 4.68818985e-01]\n",
      " [1.73654275e-04 8.25000000e-01 5.75513196e-02 ... 2.23404255e-01\n",
      "  9.96167230e-01 3.80794702e-02]\n",
      " [5.49837765e-02 0.00000000e+00 6.46627566e-01 ... 8.08510638e-01\n",
      "  9.46089061e-01 4.22185430e-02]\n",
      " ...\n",
      " [3.18534767e-04 3.50000000e-01 2.05278592e-01 ... 4.57446809e-01\n",
      "  9.12627969e-01 1.68322296e-01]\n",
      " [2.40852297e-02 0.00000000e+00 7.00879765e-01 ... 2.23404255e-01\n",
      "  6.59715568e-01 3.87969095e-01]\n",
      " [9.07048543e-05 6.00000000e-01 9.05425220e-02 ... 3.19148936e-01\n",
      "  9.49064501e-01 7.31236203e-02]]\n"
     ]
    }
   ],
   "source": [
    "# Print the normalized data\n",
    "print(train_data_normalized)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "8421f420",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[2.03195508e-01 0.00000000e+00 6.46627566e-01 ... 8.08510638e-01\n",
      "  6.79055928e-02 7.53863135e-01]\n",
      " [1.31471460e-03 0.00000000e+00 3.50073314e-01 ... 5.53191489e-01\n",
      "  9.95082959e-01 3.99558499e-01]\n",
      " [5.46814270e-04 0.00000000e+00 1.73387097e-01 ... 8.08510638e-01\n",
      "  1.00000000e+00 2.21026490e-01]\n",
      " ...\n",
      " [2.05400974e-02 0.00000000e+00 7.00879765e-01 ... 2.23404255e-01\n",
      "  9.81617832e-01 5.24282561e-03]\n",
      " [3.95380999e-03 0.00000000e+00 2.10410557e-01 ... 5.10638298e-01\n",
      "  9.86887891e-01 2.20198675e-01]\n",
      " [3.27940197e-02 0.00000000e+00 7.00879765e-01 ... 2.23404255e-01\n",
      "  6.04770790e-01 2.22958057e-01]]\n"
     ]
    }
   ],
   "source": [
    "print(test_data_normalized)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "a97b1e1a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(404, 13)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data_normalized.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d8524a22",
   "metadata": {},
   "source": [
    "## Building the Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "5919b11f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_model():\n",
    "    model = models.Sequential()\n",
    "    model.add(layers.Dense(64, activation='relu', input_shape = (train_data_normalized.shape[1],)))\n",
    "    model.add(layers.Dense(64, activation='relu'))\n",
    "    model.add(layers.Dense(1))\n",
    "    model.compile(optimizer='rmsprop', loss='mse', metrics = ['mae'])\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16cf8041",
   "metadata": {},
   "source": [
    "### K-fold CrossValidation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "9faeb9b4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the number of folds\n",
    "k = 4\n",
    "\n",
    "# Initialize lists to store the evaluation scores\n",
    "all_mae_scores = []\n",
    "\n",
    "# Perform k-fold cross-validation\n",
    "kfold = KFold(n_splits = k)\n",
    "\n",
    "for train_index, val_index in kfold.split(train_data_normalized):\n",
    "    # Split the data into training and validation sets\n",
    "    partial_train_data, val_data = train_data_normalized[train_index], train_data_normalized[val_index]\n",
    "    partial_train_targets, val_targets = train_targets[train_index], train_targets[val_index]\n",
    "\n",
    "    # Build the model\n",
    "    model = build_model()\n",
    "\n",
    "    # Train the model\n",
    "    history = model.fit(partial_train_data, partial_train_targets, epochs=100, batch_size=1, verbose=0,\n",
    "                        validation_data=(val_data, val_targets))\n",
    "\n",
    "    # Get the MAE from the history object\n",
    "    val_mae = history.history['val_mae']\n",
    "\n",
    "    # Store the evaluation score\n",
    "    all_mae_scores.append(val_mae)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "a8742bc5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean MAE: 2.828928308188915\n"
     ]
    }
   ],
   "source": [
    "# Compute the mean of the evaluation scores\n",
    "mean_mae = np.mean(all_mae_scores)\n",
    "print(\"Mean MAE:\", mean_mae)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cbf35ebb",
   "metadata": {},
   "source": [
    "## Try with 500 epochs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "5e1918b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the number of folds\n",
    "k = 4\n",
    "\n",
    "# Initialize lists to store the evaluation scores\n",
    "all_mae_histories = []\n",
    "\n",
    "# Perform k-fold cross-validation\n",
    "kfold = KFold(n_splits=k)\n",
    "for train_index, val_index in kfold.split(train_data_normalized):\n",
    "    # Split the data into training and validation sets\n",
    "    partial_train_data, val_data = train_data_normalized[train_index], train_data_normalized[val_index]\n",
    "    partial_train_targets, val_targets = train_targets[train_index], train_targets[val_index]\n",
    "\n",
    "    # Build the model\n",
    "    model = build_model()\n",
    "\n",
    "    # Train the model\n",
    "    history = model.fit(partial_train_data, partial_train_targets, epochs=500, batch_size=1, verbose=0,\n",
    "                        validation_data=(val_data, val_targets))\n",
    "\n",
    "    # Get the MAE scores for each epoch\n",
    "    val_mae_history = history.history['val_mae']\n",
    "\n",
    "    # Store the MAE scores for this fold\n",
    "    all_mae_histories.append(val_mae_history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "308f4f2a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average MAE Scores for Each Epoch:\n",
      "[6.80689299 5.3136133  4.9233247  4.53816777 4.37994432 4.15035248\n",
      " 3.8266747  3.63952219 3.74362409 3.40894288 3.29493934 3.51894474\n",
      " 3.23411298 3.23952907 3.2128588  3.14536852 3.14204377 3.05162615\n",
      " 3.20934433 2.97394449 3.09296948 2.90551281 2.94332343 2.97623324\n",
      " 2.97418725 2.83852482 2.89726365 3.03225857 2.8545078  2.63229793\n",
      " 2.72902215 2.65972924 2.60914689 2.69559652 2.87478775 2.77057934\n",
      " 2.68632072 2.53688592 2.51791799 2.58254659 2.48746204 2.54521823\n",
      " 2.57698226 2.56879419 2.58442241 2.5604412  2.66566783 2.48531801\n",
      " 2.4291653  2.63801444 2.5910877  2.65823591 2.46215492 2.42949224\n",
      " 2.63008326 2.4321391  2.43103731 2.42283016 2.5564574  2.80141515\n",
      " 2.46038604 2.46785879 2.48368257 2.36366135 2.49440151 2.3796919\n",
      " 2.43185753 2.53896594 2.35533553 2.49327129 2.3981306  2.42201465\n",
      " 2.38925248 2.36586392 2.48943555 2.30560195 2.37444973 2.34823895\n",
      " 2.33440524 2.33856422 2.40365338 2.43405861 2.41567698 2.34090418\n",
      " 2.33631355 2.39569086 2.66764849 2.30295363 2.40777814 2.28499472\n",
      " 2.20211345 2.23849189 2.24097353 2.32951659 2.3624987  2.45317322\n",
      " 2.19802111 2.28678781 2.30775172 2.25991023 2.30171692 2.27488273\n",
      " 2.33507431 2.27128386 2.26651621 2.42377341 2.19201392 2.25940531\n",
      " 2.204202   2.27553672 2.44397163 2.36192721 2.29198971 2.21787563\n",
      " 2.335262   2.31779122 2.44620723 2.16459572 2.31873685 2.15949768\n",
      " 2.30096608 2.2035965  2.60372147 2.2363354  2.18240732 2.17176017\n",
      " 2.18198669 2.1879853  2.08465707 2.18935892 2.21291366 2.16612798\n",
      " 2.25066531 2.10544181 2.12302136 2.22547776 2.20191479 2.23288634\n",
      " 2.28262308 2.23679143 2.11220458 2.10955375 2.10830039 2.25566015\n",
      " 2.11503592 2.18825018 2.46061519 2.17235494 2.3452301  2.21487015\n",
      " 2.1536319  2.12071168 2.15525872 2.1208654  2.14695257 2.19841075\n",
      " 2.19974107 2.110535   2.1764333  2.09351194 2.19279081 2.30445704\n",
      " 2.3081978  2.14581603 2.30716434 2.13080993 2.01471099 2.06116107\n",
      " 2.15484184 2.12940979 2.12877616 2.15015173 2.0249666  2.19322327\n",
      " 2.3182171  2.168414   2.05465069 2.14316154 2.12746474 2.3316524\n",
      " 2.19544104 2.1481967  2.05151498 2.21441698 2.0603438  2.06010431\n",
      " 2.05573612 2.18993556 2.30178326 2.0576179  2.28015345 2.07458422\n",
      " 2.1001493  2.17513493 2.17291832 2.23999655 2.15798703 2.08108419\n",
      " 2.11431476 2.08267802 2.06344831 2.12782353 2.14950868 2.07947829\n",
      " 2.15918338 2.08183852 2.15077275 2.08468109 2.17463237 2.21125662\n",
      " 2.08349389 1.99818662 2.32246822 2.07076833 2.09002542 2.19440234\n",
      " 2.01162741 2.11265561 2.23453707 2.01925823 2.05449247 2.06086314\n",
      " 2.00494459 2.09965071 2.16433987 2.12775743 2.34921145 2.27370375\n",
      " 2.06175676 1.98394123 2.04616696 2.13031498 2.10802633 2.12112904\n",
      " 2.46029109 1.98184136 2.08676922 2.23529807 2.10948223 2.22387692\n",
      " 2.26621187 2.20973507 2.15762794 2.02911922 1.98623428 2.20566744\n",
      " 2.13288462 2.25466534 2.10875458 2.08359334 2.17283261 2.07390222\n",
      " 2.23562765 2.04203051 2.10866481 2.10173464 2.04349822 2.08821413\n",
      " 2.11905056 1.99237359 2.09268895 2.03118193 2.15856388 2.08485341\n",
      " 2.2706933  2.01990843 2.11738828 2.11089727 2.07989874 2.11725923\n",
      " 2.03476113 2.11459422 2.17899418 2.03835139 2.34585774 2.27112216\n",
      " 2.10459611 2.04896551 2.15804395 2.00033137 2.11672187 2.03420356\n",
      " 2.19384095 2.18618122 2.06140044 2.0348554  2.02949435 2.03168473\n",
      " 2.10446921 2.17522699 2.14084202 2.18514541 2.04696929 2.02415457\n",
      " 2.21477941 2.01948345 2.10332304 2.01751417 2.02487576 2.05634475\n",
      " 2.06629878 2.15026444 2.06760824 2.10168567 2.20100525 2.20817393\n",
      " 2.02211669 2.07152212 2.03599364 2.05325967 2.09353694 2.23737651\n",
      " 2.04681933 2.04280797 2.22647619 2.11245039 2.26178652 2.25310025\n",
      " 2.11004129 2.03721938 2.04548195 2.06173739 2.07527977 2.08996546\n",
      " 2.26846078 2.20483088 2.04783857 2.14989012 2.08646914 2.1452159\n",
      " 2.11034244 2.30237287 2.3607977  2.03994057 2.06250113 2.01244637\n",
      " 2.0810234  2.12136507 2.11655539 2.13805786 2.16962013 2.35059288\n",
      " 2.20399863 2.09674954 2.08368355 2.13402259 2.15543583 2.22345737\n",
      " 2.02435079 2.10453823 2.16457412 2.1672655  2.09020561 2.20677617\n",
      " 2.13206536 2.04810932 2.07872528 2.06583363 2.12456334 2.15553135\n",
      " 2.09195018 2.23405159 2.12971294 2.30763423 2.06007388 2.25159746\n",
      " 2.11983135 2.09408858 2.11688027 2.01048282 2.14134663 2.1443052\n",
      " 2.09511164 2.10204372 2.24946567 2.06552556 2.06792858 2.12321785\n",
      " 2.15092039 2.07861003 2.08921349 2.08283105 2.16793972 2.30636263\n",
      " 2.14575827 2.17936665 2.19667146 2.17631066 2.11775029 2.22616735\n",
      " 2.05733109 2.04090768 2.211456   2.36879522 2.27137107 2.13986114\n",
      " 2.15222511 2.08477703 2.12352976 2.24234664 2.10436228 2.05112612\n",
      " 2.13400629 2.06285399 2.17766184 2.10952812 2.21256393 2.13945985\n",
      " 2.15419894 2.0295558  2.1783933  2.21004173 2.04764193 2.02237824\n",
      " 2.0585106  2.0027265  2.0527803  2.20890051 2.24521208 2.19424364\n",
      " 2.16388968 2.17248586 2.10721412 2.18407378 2.11551777 2.0971081\n",
      " 2.01355067 2.14572278 2.11812758 2.20106623 2.13971707 2.17261985\n",
      " 2.13484892 2.16582507 2.26049489 2.04370382 2.13325498 2.18213922\n",
      " 2.08226874 2.05817431 2.19406688 2.07166165 2.08671215 2.16810086\n",
      " 2.09112099 2.17230353 2.19010425 2.12783608 2.25230762 2.17999437\n",
      " 2.10249883 2.0973314  2.16736624 2.20893589 2.02341822 2.26306975\n",
      " 2.10667956 2.2395938  2.1785472  2.09008297 2.06957215 2.11680406\n",
      " 2.1357165  2.02827811 2.12853581 2.03808436 2.26400974 2.06388164\n",
      " 2.12850922 2.11479041 2.17005202 2.15305588 2.13596743 2.07495698\n",
      " 2.07138586 2.02963835 2.11219761 2.18877083 2.03526014 2.07697693\n",
      " 2.03102008 2.04257506 2.092742   2.11609319 2.21404162 2.16412055\n",
      " 2.04663244 2.12165365 2.1140106  2.20043755 2.06273866 2.23497722\n",
      " 2.14313912 2.13498977 2.13680565 2.24489984 2.12111169 2.20043159\n",
      " 2.31514364 2.0666694 ]\n"
     ]
    }
   ],
   "source": [
    "# Compute the average MAE score for each epoch across all folds\n",
    "average_mae_history = np.mean(all_mae_histories, axis=0)\n",
    "\n",
    "print(\"Average MAE Scores for Each Epoch:\")\n",
    "print(average_mae_history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "5dd9dea5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the validation MAE scores\n",
    "plt.plot(range(1, len(average_mae_history) + 1), average_mae_history)\n",
    "plt.title('Validation MAE with respect to Epoch')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Validation MAE')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6af3c4c",
   "metadata": {},
   "source": [
    "### Smothing the Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "d40fff1f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def smooth_curve(points, factor=0.9):\n",
    "    smoothed_points = []\n",
    "    for point in points:\n",
    "        if smoothed_points:\n",
    "            previous = smoothed_points[-1]\n",
    "            smoothed_points.append(previous*factor+point*(1 - factor))\n",
    "        else:\n",
    "            smoothed_points.append(point)\n",
    "    return smoothed_points\n",
    "\n",
    "smooth_mae_history = smooth_curve(average_mae_history[10:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "052dfe74",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.clf()\n",
    "plt.plot(range(1, len(smooth_mae_history) + 1), smooth_mae_history)\n",
    "plt.title('Validation MAE with respect to Epoch')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Validation MAE')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52bee54c",
   "metadata": {},
   "source": [
    "### Training the Final Model\n",
    "\n",
    "From the above graph we can see that the Validation MAE is loawest at the epoch number of around 295, so we we use the number of epochs as 295 in our final model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "ee4af96d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x22b42133550>"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = build_model()\n",
    "model.fit(train_data, train_targets, epochs= 295, batch_size=16, verbose=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "8fcf6643",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4/4 [==============================] - 0s 4ms/step - loss: 22.7490 - mae: 3.4278\n"
     ]
    }
   ],
   "source": [
    "test_mse_score, test_mae_score = model.evaluate(test_data, test_targets)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}