{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "provenance": [] }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "language_info": { "name": "python" } }, "cells": [ { "cell_type": "code", "execution_count": 1, "metadata": { "id": "ihc2awUHZidF" }, "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "import seaborn as sns\n", "import pandas as pd\n", "import numpy as np" ] }, { "cell_type": "code", "source": [ "from google.colab import drive\n", "drive.mount('/content/gdrive')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "qUekCM8Laut7", "outputId": "3eeff161-15a0-485e-eda1-7714a0029fa4" }, "execution_count": 2, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Mounted at /content/gdrive\n" ] } ] }, { "cell_type": "code", "source": [ "df = pd.read_excel('/content/gdrive/My Drive/StudyData/cancer_patient_dataset.xlsx')\n", "df.head()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 352 }, "id": "GLMNoL6jau0T", "outputId": "f4fb25e9-0706-4f17-9ee0-17f141634bc3" }, "execution_count": 3, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " Patient Id Age Gender Air Pollution Alcohol use Dust Allergy \\\n", "0 P1 33 1 2 4 5 \n", "1 P10 17 1 3 1 5 \n", "2 P100 35 1 4 5 6 \n", "3 P1000 37 1 7 7 7 \n", "4 P101 46 1 6 8 7 \n", "\n", " OccuPational Hazards Genetic Risk chronic Lung Disease Balanced Diet \\\n", "0 4 3 2 2 \n", "1 3 4 2 2 \n", "2 5 5 4 6 \n", "3 7 6 7 7 \n", "4 7 7 6 7 \n", "\n", " ... Fatigue Weight Loss Shortness of Breath Wheezing \\\n", "0 ... 3 4 2 2 \n", "1 ... 1 3 7 8 \n", "2 ... 8 7 9 2 \n", "3 ... 4 2 3 1 \n", "4 ... 3 2 4 1 \n", "\n", " Swallowing Difficulty Clubbing of Finger Nails Frequent Cold Dry Cough \\\n", "0 3 1 2 3 \n", "1 6 2 1 7 \n", "2 1 4 6 7 \n", "3 4 5 6 7 \n", "4 4 2 4 2 \n", "\n", " Snoring Level \n", "0 4 Low \n", "1 2 Medium \n", "2 2 High \n", "3 5 High \n", "4 3 High \n", "\n", "[5 rows x 25 columns]" ], "text/html": [ "\n", "
\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Patient IdAgeGenderAir PollutionAlcohol useDust AllergyOccuPational HazardsGenetic Riskchronic Lung DiseaseBalanced Diet...FatigueWeight LossShortness of BreathWheezingSwallowing DifficultyClubbing of Finger NailsFrequent ColdDry CoughSnoringLevel
0P13312454322...342231234Low
1P101713153422...137862172Medium
2P1003514565546...879214672High
3P10003717777677...423145675High
4P1014616877767...324142423High
\n", "

5 rows × 25 columns

\n", "
\n", " \n", " \n", " \n", "\n", " \n", "
\n", "
\n", " " ] }, "metadata": {}, "execution_count": 3 } ] }, { "cell_type": "code", "source": [ "df.columns" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "YIYyfrKyau7Q", "outputId": "1e5159a0-e271-42e3-f949-d43493b0a819" }, "execution_count": 4, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "Index(['Patient Id', 'Age', 'Gender', 'Air Pollution', 'Alcohol use',\n", " 'Dust Allergy', 'OccuPational Hazards', 'Genetic Risk',\n", " 'chronic Lung Disease', 'Balanced Diet', 'Obesity', 'Smoking',\n", " 'Passive Smoker', 'Chest Pain', 'Coughing of Blood', 'Fatigue',\n", " 'Weight Loss', 'Shortness of Breath', 'Wheezing',\n", " 'Swallowing Difficulty', 'Clubbing of Finger Nails', 'Frequent Cold',\n", " 'Dry Cough', 'Snoring', 'Level'],\n", " dtype='object')" ] }, "metadata": {}, "execution_count": 4 } ] }, { "cell_type": "code", "source": [ "df['Level'].unique()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "fXTL-xe6ku86", "outputId": "d3c9a8bd-429f-45f4-b731-5863b339aa81" }, "execution_count": 5, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "array(['Low', 'Medium', 'High'], dtype=object)" ] }, "metadata": {}, "execution_count": 5 } ] }, { "cell_type": "code", "source": [ "df['Level'].value_counts()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "cyYIzAzPlW7v", "outputId": "553acafb-5fd8-4ce4-8998-ff25f3f10081" }, "execution_count": 6, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "High 365\n", "Medium 332\n", "Low 303\n", "Name: Level, dtype: int64" ] }, "metadata": {}, "execution_count": 6 } ] }, { "cell_type": "code", "source": [ "df.info()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "M-spoz6WljnZ", "outputId": "f0ef8dbe-12db-46ac-871a-dba7b078b522" }, "execution_count": 7, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\n", "RangeIndex: 1000 entries, 0 to 999\n", "Data columns (total 25 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 Patient Id 1000 non-null object\n", " 1 Age 1000 non-null int64 \n", " 2 Gender 1000 non-null int64 \n", " 3 Air Pollution 1000 non-null int64 \n", " 4 Alcohol use 1000 non-null int64 \n", " 5 Dust Allergy 1000 non-null int64 \n", " 6 OccuPational Hazards 1000 non-null int64 \n", " 7 Genetic Risk 1000 non-null int64 \n", " 8 chronic Lung Disease 1000 non-null int64 \n", " 9 Balanced Diet 1000 non-null int64 \n", " 10 Obesity 1000 non-null int64 \n", " 11 Smoking 1000 non-null int64 \n", " 12 Passive Smoker 1000 non-null int64 \n", " 13 Chest Pain 1000 non-null int64 \n", " 14 Coughing of Blood 1000 non-null int64 \n", " 15 Fatigue 1000 non-null int64 \n", " 16 Weight Loss 1000 non-null int64 \n", " 17 Shortness of Breath 1000 non-null int64 \n", " 18 Wheezing 1000 non-null int64 \n", " 19 Swallowing Difficulty 1000 non-null int64 \n", " 20 Clubbing of Finger Nails 1000 non-null int64 \n", " 21 Frequent Cold 1000 non-null int64 \n", " 22 Dry Cough 1000 non-null int64 \n", " 23 Snoring 1000 non-null int64 \n", " 24 Level 1000 non-null object\n", "dtypes: int64(23), object(2)\n", "memory usage: 195.4+ KB\n" ] } ] }, { "cell_type": "code", "source": [ "risk_dictionary_binary_class = {'High':1,'Medium':0,'Low':0}\n", "df['Target Binary']=df['Level'].map(risk_dictionary_binary_class)" ], "metadata": { "id": "3Y_ctTa0l7ST" }, "execution_count": 8, "outputs": [] }, { "cell_type": "code", "source": [ "df[['Level','Target Binary']].head()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 206 }, "id": "pRrepHydmh-d", "outputId": "21a72f12-104b-46f3-8fb5-f8b5be98abe5" }, "execution_count": 9, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " Level Target Binary\n", "0 Low 0\n", "1 Medium 0\n", "2 High 1\n", "3 High 1\n", "4 High 1" ], "text/html": [ "\n", "
\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
LevelTarget Binary
0Low0
1Medium0
2High1
3High1
4High1
\n", "
\n", " \n", " \n", " \n", "\n", " \n", "
\n", "
\n", " " ] }, "metadata": {}, "execution_count": 9 } ] }, { "cell_type": "code", "source": [ "X = df.drop(['Patient Id','Level','Target Binary'],axis=1)\n", "y = df['Target Binary']" ], "metadata": { "id": "tgd-JAXjavB2" }, "execution_count": 10, "outputs": [] }, { "cell_type": "code", "source": [ "from sklearn.preprocessing import MinMaxScaler\n", "min_max_scaler = MinMaxScaler()\n", "X = min_max_scaler.fit_transform(X)" ], "metadata": { "id": "zSRjr4WFnTw6" }, "execution_count": 11, "outputs": [] }, { "cell_type": "code", "source": [ "from sklearn.model_selection import train_test_split\n", "X_train, X_test, y_train, y_test = train_test_split(X, y, stratify=y, test_size = 0.2, random_state = 0)\n", "print(f'Train : {X_train.shape}, Test : {X_test.shape}')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "iTBHmHziavHh", "outputId": "4104fb8d-6f8b-4c23-b16a-a70471cf4007" }, "execution_count": 12, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Train : (800, 23), Test : (200, 23)\n" ] } ] }, { "cell_type": "code", "source": [ "from keras.models import Sequential\n", "from keras.layers import Dense\n", "binary_classifier = Sequential()\n", "binary_classifier.add(Dense(4,activation='relu',input_dim=23))\n", "binary_classifier.add(Dense(1,activation='sigmoid'))" ], "metadata": { "id": "B3VWebzgoslT" }, "execution_count": 13, "outputs": [] }, { "cell_type": "code", "source": [ "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, stratify=y_train, test_size = 0.2, random_state = 0)\n", "print(f'Train : {X_train.shape}, Test : {X_val.shape}')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "PYm-1NdbqxmP", "outputId": "13cc0eba-438b-48c6-b18c-5647ff2855a4" }, "execution_count": 14, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Train : (640, 23), Test : (160, 23)\n" ] } ] }, { "cell_type": "code", "source": [ "binary_classifier.compile(loss='binary_crossentropy',optimizer='adam',\n", " metrics='accuracy')" ], "metadata": { "id": "cT8h0SJ0qyJq" }, "execution_count": 15, "outputs": [] }, { "cell_type": "code", "source": [ "binary_classifier_history = binary_classifier.fit(X_train, y_train, batch_size=10,\n", " validation_data=(X_val,y_val), epochs=50)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "9tELlv_RtHlf", "outputId": "722ffd79-21aa-4c72-8871-65b87d710bf4" }, "execution_count": 16, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Epoch 1/50\n", "64/64 [==============================] - 1s 6ms/step - loss: 0.6709 - accuracy: 0.5281 - val_loss: 0.6627 - val_accuracy: 0.6250\n", "Epoch 2/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.6401 - accuracy: 0.6297 - val_loss: 0.6442 - val_accuracy: 0.6375\n", "Epoch 3/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.6117 - accuracy: 0.6344 - val_loss: 0.6183 - val_accuracy: 0.6375\n", "Epoch 4/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.5718 - accuracy: 0.6344 - val_loss: 0.5799 - val_accuracy: 0.6375\n", "Epoch 5/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.5307 - accuracy: 0.6344 - val_loss: 0.5501 - val_accuracy: 0.6375\n", "Epoch 6/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.4996 - accuracy: 0.6344 - val_loss: 0.5275 - val_accuracy: 0.6375\n", "Epoch 7/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.4743 - accuracy: 0.6344 - val_loss: 0.5078 - val_accuracy: 0.6375\n", "Epoch 8/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.4521 - accuracy: 0.6344 - val_loss: 0.4873 - val_accuracy: 0.6375\n", "Epoch 9/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.4322 - accuracy: 0.6344 - val_loss: 0.4716 - val_accuracy: 0.6375\n", "Epoch 10/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.4145 - accuracy: 0.7344 - val_loss: 0.4583 - val_accuracy: 0.7312\n", "Epoch 11/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3972 - accuracy: 0.8797 - val_loss: 0.4424 - val_accuracy: 0.8062\n", "Epoch 12/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3827 - accuracy: 0.8906 - val_loss: 0.4295 - val_accuracy: 0.8188\n", "Epoch 13/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3685 - accuracy: 0.9125 - val_loss: 0.4212 - val_accuracy: 0.7937\n", "Epoch 14/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3562 - accuracy: 0.9187 - val_loss: 0.4082 - val_accuracy: 0.8250\n", "Epoch 15/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3431 - accuracy: 0.9438 - val_loss: 0.3975 - val_accuracy: 0.8562\n", "Epoch 16/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3313 - accuracy: 0.9484 - val_loss: 0.3850 - val_accuracy: 0.8687\n", "Epoch 17/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3197 - accuracy: 0.9500 - val_loss: 0.3743 - val_accuracy: 0.8687\n", "Epoch 18/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3096 - accuracy: 0.9500 - val_loss: 0.3630 - val_accuracy: 0.8687\n", "Epoch 19/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2977 - accuracy: 0.9500 - val_loss: 0.3501 - val_accuracy: 0.8687\n", "Epoch 20/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.2872 - accuracy: 0.9516 - val_loss: 0.3406 - val_accuracy: 0.8813\n", "Epoch 21/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.2773 - accuracy: 0.9500 - val_loss: 0.3249 - val_accuracy: 0.8687\n", "Epoch 22/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2663 - accuracy: 0.9563 - val_loss: 0.3132 - val_accuracy: 0.8687\n", "Epoch 23/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2557 - accuracy: 0.9547 - val_loss: 0.2993 - val_accuracy: 0.9375\n", "Epoch 24/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2457 - accuracy: 0.9703 - val_loss: 0.2865 - val_accuracy: 0.9375\n", "Epoch 25/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2361 - accuracy: 0.9734 - val_loss: 0.2800 - val_accuracy: 0.9438\n", "Epoch 26/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.2276 - accuracy: 0.9703 - val_loss: 0.2647 - val_accuracy: 0.9438\n", "Epoch 27/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2191 - accuracy: 0.9719 - val_loss: 0.2542 - val_accuracy: 0.9688\n", "Epoch 28/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2111 - accuracy: 0.9750 - val_loss: 0.2437 - val_accuracy: 0.9688\n", "Epoch 29/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2035 - accuracy: 0.9859 - val_loss: 0.2340 - val_accuracy: 0.9688\n", "Epoch 30/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1966 - accuracy: 0.9875 - val_loss: 0.2250 - val_accuracy: 0.9688\n", "Epoch 31/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1894 - accuracy: 0.9922 - val_loss: 0.2168 - val_accuracy: 0.9937\n", "Epoch 32/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1831 - accuracy: 0.9953 - val_loss: 0.2099 - val_accuracy: 0.9937\n", "Epoch 33/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1764 - accuracy: 0.9937 - val_loss: 0.2009 - val_accuracy: 0.9937\n", "Epoch 34/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1705 - accuracy: 0.9937 - val_loss: 0.1929 - val_accuracy: 0.9937\n", "Epoch 35/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1645 - accuracy: 0.9953 - val_loss: 0.1863 - val_accuracy: 0.9937\n", "Epoch 36/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1595 - accuracy: 0.9937 - val_loss: 0.1794 - val_accuracy: 0.9937\n", "Epoch 37/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1536 - accuracy: 0.9953 - val_loss: 0.1773 - val_accuracy: 0.9875\n", "Epoch 38/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1486 - accuracy: 1.0000 - val_loss: 0.1659 - val_accuracy: 1.0000\n", "Epoch 39/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1446 - accuracy: 0.9969 - val_loss: 0.1608 - val_accuracy: 0.9937\n", "Epoch 40/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1392 - accuracy: 0.9984 - val_loss: 0.1544 - val_accuracy: 1.0000\n", "Epoch 41/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1350 - accuracy: 1.0000 - val_loss: 0.1494 - val_accuracy: 1.0000\n", "Epoch 42/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1306 - accuracy: 1.0000 - val_loss: 0.1442 - val_accuracy: 1.0000\n", "Epoch 43/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1269 - accuracy: 1.0000 - val_loss: 0.1411 - val_accuracy: 1.0000\n", "Epoch 44/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1227 - accuracy: 1.0000 - val_loss: 0.1358 - val_accuracy: 1.0000\n", "Epoch 45/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1187 - accuracy: 1.0000 - val_loss: 0.1308 - val_accuracy: 1.0000\n", "Epoch 46/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1152 - accuracy: 1.0000 - val_loss: 0.1289 - val_accuracy: 1.0000\n", "Epoch 47/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1117 - accuracy: 1.0000 - val_loss: 0.1242 - val_accuracy: 1.0000\n", "Epoch 48/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1083 - accuracy: 1.0000 - val_loss: 0.1181 - val_accuracy: 1.0000\n", "Epoch 49/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1047 - accuracy: 1.0000 - val_loss: 0.1156 - val_accuracy: 1.0000\n", "Epoch 50/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1017 - accuracy: 1.0000 - val_loss: 0.1112 - val_accuracy: 1.0000\n" ] } ] }, { "cell_type": "code", "source": [ "loss_function = binary_classifier_history.history['loss']\n", "val_loss_function = binary_classifier_history.history['val_loss']\n", "epochs = range(1,len(loss_function)+1)\n", "\n", "plt.title('Loss function (Train & Val Sets)')\n", "plt.plot(epochs,loss_function,label='Train Loss')\n", "plt.plot(epochs,val_loss_function,color='orange',label='Val Loss')\n", "plt.xlabel('Epochs')\n", "plt.ylabel('Loss function')\n", "plt.legend()\n", "plt.show()" ], "metadata": { "id": "WibrpHvXb_ET", "colab": { "base_uri": "https://localhost:8080/", "height": 295 }, "outputId": "3f7211eb-547a-4b5b-8988-9bcaf36e0dfe" }, "execution_count": 17, "outputs": [ { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": { "needs_background": "light" } } ] }, { "cell_type": "code", "source": [ "acc = binary_classifier_history.history['accuracy']\n", "val_acc = binary_classifier_history.history['val_accuracy']\n", "epochs = range(1,len(acc)+1)\n", "\n", "plt.title('Accuracy (Train & Val Sets)')\n", "plt.plot(epochs,acc,label='Accuracy (Train)')\n", "plt.plot(epochs,val_acc,color='orange',label='Accuracy (Validation)')\n", "plt.xlabel('Epochs')\n", "plt.ylabel('Accuracy')\n", "plt.legend()\n", "plt.show()" ], "metadata": { "id": "V3xHmBeRb_JW", "colab": { "base_uri": "https://localhost:8080/", "height": 295 }, "outputId": "fa7a27cc-d584-431b-ad5e-3cd19a6074c2" }, "execution_count": 18, "outputs": [ { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "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\n" }, "metadata": { "needs_background": "light" } } ] }, { "cell_type": "code", "source": [ "results = binary_classifier.evaluate(X_test,y_test)" ], "metadata": { "id": "47tntoU5b_N5", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "16404969-ee71-4cb9-d11d-88836e3682cb" }, "execution_count": 19, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "7/7 [==============================] - 0s 2ms/step - loss: 0.1081 - accuracy: 1.0000\n" ] } ] }, { "cell_type": "code", "source": [ "x_test_pattern = X_test[7,:]\n", "y_pred = binary_classifier.predict(x_test_pattern.reshape(1,-1))\n", "print(y_pred[0])" ], "metadata": { "id": "Tralp5Zbb_T7", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "31063b49-1489-4034-f96f-882589b1e03f" }, "execution_count": 20, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "1/1 [==============================] - 0s 83ms/step\n", "[0.8245223]\n" ] } ] }, { "cell_type": "code", "source": [ "print(x_test_pattern)" ], "metadata": { "id": "LBSFDD37b_YZ", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "9aeea79f-4342-4c83-85ff-83bb88e3dfe2" }, "execution_count": 21, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "[0.38983051 0. 1. 1. 0.85714286 0.85714286\n", " 1. 0.83333333 1. 1. 1. 0.85714286\n", " 0.75 1. 0.25 0.14285714 0.375 0.\n", " 0.42857143 0.125 0.5 0.16666667 0.33333333]\n" ] } ] }, { "cell_type": "code", "source": [ "original_features= min_max_scaler.inverse_transform(x_test_pattern.reshape(1,-1))\n", "original_features" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "r2SGjBik1FBf", "outputId": "89ebf879-aa55-484b-e341-d2f86e00e7db" }, "execution_count": 22, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "array([[37., 1., 8., 8., 7., 7., 7., 6., 7., 7., 8., 7., 7.,\n", " 9., 3., 2., 4., 1., 4., 2., 4., 2., 3.]])" ] }, "metadata": {}, "execution_count": 22 } ] }, { "cell_type": "code", "source": [ "risk_dictionary_multi_class = {'High':2,'Medium':1,'Low':0}\n", "df['Target Multi']=df['Level'].map(risk_dictionary_multi_class)" ], "metadata": { "id": "OnZUkdN01_LT" }, "execution_count": 23, "outputs": [] }, { "cell_type": "code", "source": [ "X = df.drop(['Patient Id','Level','Target Binary','Target Multi'],axis=1)\n", "y = df['Target Multi']" ], "metadata": { "id": "CFADVdAK1_dF" }, "execution_count": 24, "outputs": [] }, { "cell_type": "code", "source": [ "X = min_max_scaler.fit_transform(X)" ], "metadata": { "id": "Pf-5iHao1_kQ" }, "execution_count": 25, "outputs": [] }, { "cell_type": "code", "source": [ "from sklearn.model_selection import train_test_split\n", "X_all_train, X_test, y_all_train, y_test = train_test_split(X, y, stratify=y, test_size = 0.2, random_state = 0)\n", "print(f'Train : {X_all_train.shape}, Test : {X_test.shape}')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "s3722X001_tN", "outputId": "58f3686b-90bb-4ae3-d620-487e2d1fd6cb" }, "execution_count": 26, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Train : (800, 23), Test : (200, 23)\n" ] } ] }, { "cell_type": "code", "source": [ "from sklearn.model_selection import train_test_split\n", "X_train, X_val, y_train, y_val = train_test_split(X_all_train, y_all_train, \n", " stratify=y_all_train, test_size = 0.2, \n", " random_state = 0)\n", "print(f'Train : {X_train.shape}, Test : {X_val.shape}')" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "ZiYlGutE1_2C", "outputId": "34cbde0e-8930-467b-be0c-a8008aa17bb3" }, "execution_count": 27, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Train : (640, 23), Test : (160, 23)\n" ] } ] }, { "cell_type": "code", "source": [ "from tensorflow.keras.utils import to_categorical\n", "y_train = to_categorical(y_train)\n", "y_val = to_categorical(y_val)\n", "y_test = to_categorical(y_test)" ], "metadata": { "id": "Qp7YQP1c1_-G" }, "execution_count": 28, "outputs": [] }, { "cell_type": "code", "source": [ "multi_classifier = Sequential()\n", "multi_classifier.add(Dense(8,activation='relu',input_dim=23))\n", "multi_classifier.add(Dense(3,activation='softmax'))" ], "metadata": { "id": "BHETHNVz2AT6" }, "execution_count": 29, "outputs": [] }, { "cell_type": "code", "source": [ "multi_classifier.compile(loss='categorical_crossentropy',optimizer='adam',\n", " metrics='accuracy')" ], "metadata": { "id": "OjmGrZSD2Adj" }, "execution_count": 30, "outputs": [] }, { "cell_type": "code", "source": [ "multi_classifier_history = multi_classifier.fit(X_train, y_train, batch_size=10,\n", " validation_data=(X_val,y_val), epochs=50)" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "8ddO_Ut22Ahw", "outputId": "29aee2f1-0ead-4707-949b-6c91cb0e1c35" }, "execution_count": 31, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Epoch 1/50\n", "64/64 [==============================] - 1s 5ms/step - loss: 1.0349 - accuracy: 0.3766 - val_loss: 0.9720 - val_accuracy: 0.4125\n", "Epoch 2/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.9379 - accuracy: 0.4094 - val_loss: 0.9087 - val_accuracy: 0.5063\n", "Epoch 3/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.8792 - accuracy: 0.5750 - val_loss: 0.8570 - val_accuracy: 0.6750\n", "Epoch 4/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.8235 - accuracy: 0.6719 - val_loss: 0.8033 - val_accuracy: 0.7312\n", "Epoch 5/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.7674 - accuracy: 0.6953 - val_loss: 0.7530 - val_accuracy: 0.7312\n", "Epoch 6/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.7114 - accuracy: 0.7344 - val_loss: 0.6983 - val_accuracy: 0.7875\n", "Epoch 7/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.6534 - accuracy: 0.7719 - val_loss: 0.6472 - val_accuracy: 0.7750\n", "Epoch 8/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.5892 - accuracy: 0.8094 - val_loss: 0.5932 - val_accuracy: 0.8250\n", "Epoch 9/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.5336 - accuracy: 0.8422 - val_loss: 0.5499 - val_accuracy: 0.8313\n", "Epoch 10/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.4850 - accuracy: 0.8516 - val_loss: 0.5135 - val_accuracy: 0.8313\n", "Epoch 11/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.4439 - accuracy: 0.8594 - val_loss: 0.4818 - val_accuracy: 0.8125\n", "Epoch 12/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.4128 - accuracy: 0.8625 - val_loss: 0.4588 - val_accuracy: 0.8375\n", "Epoch 13/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3844 - accuracy: 0.8641 - val_loss: 0.4379 - val_accuracy: 0.8750\n", "Epoch 14/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.3598 - accuracy: 0.8766 - val_loss: 0.4172 - val_accuracy: 0.8687\n", "Epoch 15/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3378 - accuracy: 0.8750 - val_loss: 0.3996 - val_accuracy: 0.8625\n", "Epoch 16/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.3199 - accuracy: 0.8906 - val_loss: 0.3862 - val_accuracy: 0.8875\n", "Epoch 17/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.3027 - accuracy: 0.8922 - val_loss: 0.3751 - val_accuracy: 0.9000\n", "Epoch 18/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2895 - accuracy: 0.9047 - val_loss: 0.3590 - val_accuracy: 0.8875\n", "Epoch 19/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2749 - accuracy: 0.9172 - val_loss: 0.3469 - val_accuracy: 0.8875\n", "Epoch 20/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2638 - accuracy: 0.9266 - val_loss: 0.3337 - val_accuracy: 0.8938\n", "Epoch 21/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2533 - accuracy: 0.9203 - val_loss: 0.3255 - val_accuracy: 0.9250\n", "Epoch 22/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.2418 - accuracy: 0.9266 - val_loss: 0.3170 - val_accuracy: 0.9250\n", "Epoch 23/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.2322 - accuracy: 0.9406 - val_loss: 0.3089 - val_accuracy: 0.9250\n", "Epoch 24/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.2235 - accuracy: 0.9422 - val_loss: 0.3016 - val_accuracy: 0.9312\n", "Epoch 25/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2166 - accuracy: 0.9469 - val_loss: 0.2956 - val_accuracy: 0.9312\n", "Epoch 26/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2083 - accuracy: 0.9609 - val_loss: 0.2799 - val_accuracy: 0.9250\n", "Epoch 27/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.2007 - accuracy: 0.9594 - val_loss: 0.2723 - val_accuracy: 0.9250\n", "Epoch 28/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1937 - accuracy: 0.9516 - val_loss: 0.2677 - val_accuracy: 0.9312\n", "Epoch 29/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1871 - accuracy: 0.9578 - val_loss: 0.2573 - val_accuracy: 0.9312\n", "Epoch 30/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1803 - accuracy: 0.9656 - val_loss: 0.2506 - val_accuracy: 0.9312\n", "Epoch 31/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1746 - accuracy: 0.9656 - val_loss: 0.2471 - val_accuracy: 0.9312\n", "Epoch 32/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1694 - accuracy: 0.9672 - val_loss: 0.2378 - val_accuracy: 0.9312\n", "Epoch 33/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1645 - accuracy: 0.9656 - val_loss: 0.2312 - val_accuracy: 0.9312\n", "Epoch 34/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1580 - accuracy: 0.9656 - val_loss: 0.2256 - val_accuracy: 0.9312\n", "Epoch 35/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1532 - accuracy: 0.9656 - val_loss: 0.2170 - val_accuracy: 0.9312\n", "Epoch 36/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1491 - accuracy: 0.9672 - val_loss: 0.2119 - val_accuracy: 0.9563\n", "Epoch 37/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1439 - accuracy: 0.9703 - val_loss: 0.2058 - val_accuracy: 0.9563\n", "Epoch 38/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1406 - accuracy: 0.9688 - val_loss: 0.2009 - val_accuracy: 0.9563\n", "Epoch 39/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1350 - accuracy: 0.9703 - val_loss: 0.1950 - val_accuracy: 0.9563\n", "Epoch 40/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1314 - accuracy: 0.9719 - val_loss: 0.1887 - val_accuracy: 0.9563\n", "Epoch 41/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1265 - accuracy: 0.9719 - val_loss: 0.1875 - val_accuracy: 0.9563\n", "Epoch 42/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1228 - accuracy: 0.9719 - val_loss: 0.1774 - val_accuracy: 0.9563\n", "Epoch 43/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1195 - accuracy: 0.9719 - val_loss: 0.1726 - val_accuracy: 0.9563\n", "Epoch 44/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1152 - accuracy: 0.9719 - val_loss: 0.1687 - val_accuracy: 0.9563\n", "Epoch 45/50\n", "64/64 [==============================] - 0s 2ms/step - loss: 0.1121 - accuracy: 0.9719 - val_loss: 0.1645 - val_accuracy: 0.9563\n", "Epoch 46/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1088 - accuracy: 0.9719 - val_loss: 0.1596 - val_accuracy: 0.9563\n", "Epoch 47/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1046 - accuracy: 0.9734 - val_loss: 0.1537 - val_accuracy: 0.9563\n", "Epoch 48/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.1018 - accuracy: 0.9750 - val_loss: 0.1489 - val_accuracy: 0.9563\n", "Epoch 49/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.0980 - accuracy: 0.9812 - val_loss: 0.1472 - val_accuracy: 0.9563\n", "Epoch 50/50\n", "64/64 [==============================] - 0s 3ms/step - loss: 0.0956 - accuracy: 0.9766 - val_loss: 0.1398 - val_accuracy: 0.9563\n" ] } ] }, { "cell_type": "code", "source": [ "acc = multi_classifier_history.history['accuracy']\n", "val_acc = multi_classifier_history.history['val_accuracy']\n", "epochs = range(1,len(acc)+1)\n", "\n", "plt.title('Accuracy (Train & Val Sets)')\n", "plt.plot(epochs,acc,label='Accuracy (Train)')\n", "plt.plot(epochs,val_acc,color='orange',label='Accuracy (Validation)')\n", "plt.xlabel('Epochs')\n", "plt.ylabel('Accuracy')\n", "plt.legend()\n", "plt.show()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 295 }, "id": "H6bKfVsY2AmX", "outputId": "cc06c99a-8e5f-4f7c-cc40-c3051da7ecfc" }, "execution_count": 32, "outputs": [ { "output_type": "display_data", "data": { "text/plain": [ "
" ], "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd3xV9f348dc7i5DNCDNAEAHDFiIoqOAErUoVq+DEr+Kosz9tq36/X2f7rV1u2mprRa2CgoWixYXEgVYhgGyEyAgZQEhY2ev9++OchEvIuEBubnLv+/l43EfuGffc9wnhvM/5TFFVjDHGBK8QfwdgjDHGvywRGGNMkLNEYIwxQc4SgTHGBDlLBMYYE+QsERhjTJCzRGCCkoi0E5ENItK9mY/7sIj8rTmP6Ssi8piI/MPPMQwTka/9GYOxRGDqISKficg+EWnn71h86FbgC1XNFZEPRKTQfVWISLnH8l+O5aCq+n+qesvxBiUiF4rIJhE5JCKrReTURvY9XUSKRCSmnm2rROSu443DPcbDIrLN/T1kicjbXn5uuogs9WZfVV0D7BeRS08kVnNiLBGYI4hIMnAWoMBlLfzdYS34dbcDbwCo6kWqGqOqMcCbwO9qllX19haO7zXgj0AccA2wr6EdVfUbIAu40nO9iAwBBgGzjzcIEbkRuB443/29pAKfHu/xmvAmcJuPjm28YInA1HUD8A0wC7jRc4OI9BKRf4pInojki8iLHttmiMhG9052g4iMdNeriJzssd8sEfmV+36Ce6f5SxHZBbwqIh1E5H33O/a575M8Pt9RRF4VkRx3+wJ3/TrPu0oRCReRvfXdUYtIb+Ak4Numfhlu/HeKyBZgi7vuORHZKSIHRWSFiJzlsX9tcYuIJLufv1FEMt14/ruJr6wAtqtjvapub2L/13D+zTzdACxS1fzGYm3CacBHqvoDgKruUtWXPc4zXkReEZFcEckWkV+JSKiIpAB/Ac5wnyT2u/tf7P5dHHL3f8Djuz4DzgvwJ9BWzRKBqesGnDu0N4GJItIVQERCgfeBHUAy0BOY4277CfCY+9k4nCeJfC+/rxvQEeiDU1wTArzqLvcGSoAXPfZ/A4gCBgNdgGfc9a8D13nsdzGQq6qr6vnOocBWVa30MsYfA2Nw7rIBlgMj3LjfAuaKSGQjnz8TGAicBzziXiyPIiICLAP+5j6ZeeMN4GwR6eUeIwTnSeK144y1xjfADSLycxFJdf/9Pc0CKoGTgVOBC4FbVHUjztPWf9wnqgR3/1eA21Q1FhgCLKk5kKpm4yTAgV6es2luqmove6Gq4FywKoDO7vIm4Gfu+zOAPCCsns99BNzbwDEVONljeRbwK/f9BKAciGwkphHAPvd9d6Aa6FDPfj2AQ0CcuzwP+EUDx7wW+KaBbbXxecR/bhO/t33AcPf9Y8A/3PfJ7ueTPPZdBkxt4DgPAovc+H4Akt31twDvNvL9i4GH3fcXuP9O4ccSayO/p8VAEU5i/6W7vitQBrT32HcakOa+nw4srXOsTJzin7gGvisbONvf/weC9WVPBMbTjcDHqrrXXX6Lw8VDvYAdWv9ddC+cC9fxyFPV0poFEYkSkZdEZIeIHAS+ABLcO9JeQIGqHlVurqo5wFfAFBFJAC7Ceaqpzz4g9hhi3Om5ICIPuMVgB9yij3igcyOf3+Xxvhg4qnLXdS/wpKq+CfweSHOfDMbhcQddj9dwyvNxf85R1YrjjLWWqr6pqucDCTh3+U+KyEScp7VwIFdE9rvHfQnnCa0hU3Ce0naIyOcickad7bHAfm/iMs2vJSvnTCsmIu2Bq4BQt7weoB3ORXg4zsWwt4iE1ZMMdgL9Gjh0MU5RTo1uOBWcNeoOf3s/ThHBGFXdJSIjgFWAuN/TUUQSVLW+i8ZrOHfPYThFE9kNxLQG6NvAudSnNka3jP0XOMU861W1WkT2ufGdqDCcCyyq+hcR6YhTfg5wTyOf+yfwJxE5B7gC50mr2WJ1k8pcEfklTrHOWzhPBJ0b+P0dNaSxqi4HJotIOHAX8A5OYkdEegIRwPfHEpdpPvZEYGr8GKjCKQcf4b5SgC9xyv6XAbnAUyISLSKRIjLO/ezfgAdEZJQ4ThaRPu6274Br3IrEScD4JuKIxakX2O9eCB+t2aCqucAHOBe9Dm6F8Nken10AjMS5s369oS9Q1SwgAxjdRCwNxVeJW0wmIo/g1Is0h7nA70XkJHFaKC3DKdsvA+qW0ddS1SKcorBXcZ7a0k80VnGagP5IRGJFJERELsKpl/nW/Xf4GPijiMS52/uJSM2/7W4gSUQi3GNFiMi1IhLvJpWDOEV8NcYDS1S1zJvYTPOzRGBq3Ai8qqqZ6rQQ2aWqu3Aqaq/FuYu8FKdyMBPnrv5qAFWdC/wa507xEM4FuaN73Hvdz+13j7OgiTieBdoDe3EqLD+ss/16nHqMTcAe4L6aDapaArwL9MW5S27MSxwuTjkWH7kxbcapOC+lTtHRCbgfJ/F+gfP7egy4HFgN/NO9m27IazhFNp4J8ERiPQg8jPNvvR/4HXCHqtb0D7gB5y5+A05R2zycOhxwirHWA7tEpKaY8Xpgu1vcdzvO30KNa3FaGhk/EVWbmMYEDveud4CqXtfEfu1wipzOc+9wjR+IyDDgJVWtW2dgWpAlAhMw3KKkVcD1qvqFv+Mxpq2woiETEERkBk6xxweWBIw5NvZEYIwxQc6eCIwxJsi1uX4EnTt31uTkZH+HYYwxbcqKFSv2qmpifdt8lghE5O/AJcAeVR1Sz3YBnsPpbVgMTFfVlU0dNzk5mfT09KZ2M8YY40FEdjS0zZdFQ7OASY1svwjo775uBf7sw1iMMcY0wGeJwG25UdDILpOB19XxDc5QBs06W5Qxxpim+bOyuCdH9nLMctcdRURuFZF0EUnPy8trkeCMMSZYtIlWQ6r6sqqmqmpqYmK9dR3GGGOOkz8TQTbu6IOuJHedMcaYFuTPRLAQZwYkEZHTgQM25osxxrQ8XzYfnY0zLnpnEcnCGU64dqx1nJmYLsYZDrgYuMlXsRhjjGmYzxKBqk5rYrsCd/rq+40xpi2pqlZW7NjHqsx9pHSPY1SfDkS3a5k+v22uZ7ExxgSKqmpl2bYCPliXywfrdpF36PDcPGEhwtCkeMb07cSYkzqS2qcDsZGNTUlx/CwRGGOMD1VXK/tLKsgvLCO/qJyConLyi8rZmHuQj9fvYm9hOZHhIZwzsAsXDe3O6X07snHXIb7dms+32wp4ZelW/vL5D4QIPDF5CNed3qfpLz1GlgiMMa1WaUWVc/EsLCe/qIyConL2F1dQ3QpHTT5YWkmBG2N+oXPBLygqZ19xOdX1hNs+PJRzU7pw8ZDunHNKIlERhy/HXeIiGT/AaSpfUl7Fysx9fLs1nxG9EnwSuyUCY0yzOlhaQcaeQjJ2F7JlzyG27Clky+5CqqqVjtERdIqJoFN0BB2j29EpJoKwEKm9S3YuoofvnIvLq/x9Ol4TgYT24c45RrejX2IMp/WtOdeI2vU1v4OO0RGEhzbdcLN9RCjjTu7MuJM7+yx2SwTGtEG7D5by+fd5VNZzq1mtyoGSCveu9PBFdV9ROWGhIe4Fyb0wxbSjU3QEUe1CEcTr76+qrmZfcYXHBbyM/MJy9haWs7fwcDl3u7AQ+iXGMKpPByLCQmr3355fREFhOUXuhT4iLOSIC2bfztG1ieLwRdT52SEqgtBQ72NtKVHhoYR5cWFvjSwRGNNGqCpf/5DPP77ZwccbdlNVX3mDh+iIUDrGOHfeXeMiOaVbHBVV1RQUlZNzoJR1OQcoKCqnour4i1liI8NqL9BJHaIYnpRAn85RDOgSS/+uMSR1iCI0pOGLdmlFFZXVSnREKM6AxMYfLBEY08y+ytjLUx9sAjjiTrZjTASdo9sx5qSO9OkU7fXxDhRXMG9lFm9+u4OteUUkRIVz85l9mTIyiYSoo1uRCBDXPpzI8NAmj62qHCqrpOQYi2CcYpAIIsJO7A7YmxiN71kiMKaZVFZV8+ziLcz8LIM+HaPo0ymagqJyMvYUkl9URmlFde2+Zw9I5LoxvTn3lC71FieUlFfx+eY9LFq7i4837KK0oppTeyfw9FXDuXho92a7gIoIcZHhxDVXs8S930JxZvMcyxytw0iI7dfsh7VEYEwzyD1Qwj2zV7F8+z5+MiqJxycPPqIVCEBxeSW7DpSycHUOs5dlcusbK+gRH8m00b25enQvoiPCSPt+Dx+s3cWSTXsoqaiiY3QEU0Ymcc2Y3gzuEe+ns/PSzn/Cl1P8HUVgO+3PPkkEbW7y+tTUVLUZykxrsnjDbh6Yt5qKymp+fflQfnxqvaOpH6GiqppPN+7mH99ksjRjL2EhQmiIUFZZTeeYdkwa0pWLh3RndN+ObaMC8tAP8OFIiB0Ip/8djqHi2RyD9t2hXcfj+qiIrFDV1Pq22ROBMcdpb2EZf/7sB15Zuo1B3eN48ZpTOSkxxqvPhoeGMGlIdyYN6c7WvELeTt9JeWU1kwZ3IzW5Y6MVrK1OVSks/QlIKJz5DsQk+zsic4wsERjjpT2HSvl2awHfuD0+M/YUAnDjGX146OKU4y63PykxhocuSmnOUFvWivtg3yoY/54lgTbKEoExrrxDZWzIPVjbJr6mZ+jewnK25hWydW8R4DTLTE3uyBUje3LWyYkMTWrlZfe+tO1NyHgJUn4BPS/xdzTmOFkiMAZYuDqHh/+5lsKyytp1YSFCB7f550mJ0Vx9Wi9OP6kTg3vEtY1ye187sBGW3waJZ8HwX/s7GnMCLBGYoFZSXsUT769n9rKdjOrTgZ9PHEiX2HZ0im5HXPsw6+TUkMoiWHolhEbBuDkQYpeStsz+9UzQ2rL7EHe+tZItewr56YR+/OyCAfWP/VJZAkXb6z9IWDRE9z7xYMoPQEnOiR+npaz/tfNEcO7HENXD39GYE2SJwAQdVWVuehaPLFxHTLswXrtpNGe7Iz3WqiyCnEWQOQ9y/u0sNyR+EPS6Enr/BOIHO91uvVFWAFkLnO/YvRiqK47/pPxh6OPQ7Xx/R2GagSUC02ZVVysfb9jNih0FnNk/kbH9Oh15R19WAJWFtYuVVdWsytzP/O+y+ez7PVzQpwOPXtqPzjElUJQJWg17v4Gd85wkUFUC7RIh+TqnHLy+4o+SXZA1H9Y9CeuegLiBTlLodbnz2aOCroDdS9yL/xLQSohOhoH3QodR3icRf4voYEkggFiHMtPmVFcrH6zbxQtLtrBp1yFCBKoV4tuHc8Ggrlw8tBtnxqwiYulkqC4/9i+I7Aa9pkDvK90E4EWz0JJdh+/u96Q5SaUxMf2cJ4jeVzrDBrSVBGDarMY6lFkiMG1GVbXy77W5vPDpFrbsKaRfYjR3n9ufCwd35euMfBaty+WTDbspLyvm44F30S4slD/n/YTi8koiw0IZ1COWoT0TGNgttuFx4GMHQOczvLv4N6Q0D3Ytdp4o6tNxJCQMt4u/aVHWs9i0eR+t38VvP9zE1rwiBnSN4YVpp3Lx0O61PXDPH9SV8wd1pbyymuwvH6JPbg737/0D2vM8Lh7SnTP7d265kS4jEyF5Wst8lzHNwBKBadX2Fpbx6L/W8++1uQzsGsufrh3JpMHdCGlgCIaIkm303f0c9L6aP15zfwtHa0zbZInAtEqqysLVOTy2cD1FZVX8fOJAbj37pMan9lOFFfdASASMfLrlgjWmjbNEYFqd3QdL+e/561i8cTcjeiXw+yuH0b9rbNMfzFrgtPYZ+bS1bTfmGPg0EYjIJOA5IBT4m6o+VWd7H+DvQCJQAFynqlm+jMm0XqrKvBVZPPn+Bsoqq/nvi1P4rzP7ejcSZ2URrLgXEobBgLt9H6wxAcRniUBEQoGZwAVAFrBcRBaq6gaP3f4AvK6qr4nIucBvgOt9FZNpvbL3l/DwP9fy+eY8TkvuwG+nDPN6SGfAacdfvBPGzbbhDow5Rr78HzMayFDVrQAiMgeYDHgmgkHA/3PfpwELfBiPaS12/hO+exB6XooOfYK3Vu7lN4s2Ua3K45cN5vrT+zRYGVyvAxtg4x/hpJsgcZzv4jYmQPlyCMWewE6P5Sx3nafVwBXu+8uBWBHpVPdAInKriKSLSHpeXp5PgjUtoHQPLL3Kmc6wugI2Pc3utwey8OPZDEuK56P7zubGscnHlgRUYfmdEB4LI37ru9iNCWD+Hkv3AWC8iKwCxgPZQFXdnVT1ZVVNVdXUxMR6uu2bVk2rqynZ8gZV76VQvfNfZHR/mD/F/JsbdzxFRWU1b/d7iDdHvEGv2CZ643qqOATbZ8MXk2HPZzD8N077fWPMMfNl0VA20MtjOcldV0tVc3CfCEQkBpiiqvt9GJNpQR+uy+WlD7/izpg/cn7cN6wqGsjPs35FxurewDbGDziHsEtuhh2/Rr5/zhncbczfGh7DpvwAZC90hnHI/Qiqy6B9Dxj0IPS7pUXPzZhA4stEsBzoLyJ9cRLAVOAazx1EpDNQoKrVwEM4LYhMG6eqPP9pBp9+9T5v9XuEdiEV/Cf+IbIH3MZ/n9WejtERdI5tR4/4SGe8/87POOPufPtfsOQCiO4LUvdhVZ3K4OoKiEqC/nc44/R0PqOefY0xx8JniUBVK0XkLuAjnOajf1fV9SLyBJCuqguBCcBvRESBL4A7fRWPaRnF5ZX8fO4alm7YTNrg3xMd1Qk59xPOiOvf+AcTx8JF3zmVvgc31r9P+ynQ6wroNNou/sY0Ixt0zjSb7P0lzHgtnU279pN22jP0Ll+KnL8UOo/2d2jGBD0bdM40iz0HS/khr4iO0RF0jI6gQ1R47dy96dsLuP0fKyirqOaTicvok5MGo16wJGBMG2CJwBxJ1emhW13ulMF3mUA1obzxzQ5+++EmissPN+oSceYA6BgdQVZBCT0SIlnwkzKSVv7GKfMfYCV9xrQFlgjMkXI/hs0vgIRBxktUhXfks6JxfJqVypjk85l+Zn8OlVZQUFROfmE5BUXOa0zfjjw4oQPxn4+BmJOc1j823r4xbYIlAnOYKqx9FKL7UDXxOz79fA5lW9/mnJhPOe+k99CIZ5GCyc5UjIPPh9B2hz9bXQVpE6F8H0z4EMLj/HcexphjYonA1KrM+jdh+d+yvd+z3PfKer7b2YvzU37H6Ev6E1P0JZI5F3bOh62zIDweel7mFB91vxDW/wZ2fwpjXoEOw/x9KsaYY2CJIIiUVlTxQ14hGXsK2bK7kB/yCsk7VOYU8xSV8UbPn5EQ2pXz5ycTG1XEc1NHcNnwHk5b/w6XQtKlUFXuXPAz5zrDPm9/A8JinNE/T5oO/f7L36dpjDlGlggCWGlFFR+sy2XR2l1s3n2IzIJialoLh4YIfTpG0TUukpQecYxt9xXDyjL4uvPveG7UaM7o14mO0RFHHzQ0Anpc5LyqX4JdS2DnPCgvgNSZLXuCxphmYf0IAlBmfjFvLtvB3PQsCorK6ZnQnhG9Eji5Swz9u8bQv0ssyZ2jaBfmzuGrCh+OdMbvuWSTDeNsTACyfgRBoKpaSdu0hze+2cEXW/IIEeH8lC5cd3ofxvXr3PiInlkLYN93cMbrlgSMCUL2v76Nq6yq5l/f5TAzLYOte4voEtuOu8/tz7TRvege377pA2i101IodgD0meb7gI0xrY4lgjaqoqqa+auymZmWwY78Yk7pFsuL15zKxMHdGp/gva6d78L+tTD2TXsaMCZI2f/8Nqa8spp/rsxi5mcZ7CwoYXCPOF66fhQXpHQ9tgldwGn7v/YxiEuB3lf7JF5jTOtniaANUVWmv7qMr3/IZ1hSPI9dOphzT+niNO88HplznWkex82BkNDmDdYY02ZYImhD3luTy9c/5PM/P0rh5jP7Hn8CAKiuhHWPQ/xgZ1wgY0zQskTQRhSXV/KbRRsZ0jOOm8adYBIAWPO/cHATnDXfxvY3JshZImgj/vL5Vg4dKuAf5+8ndOPS+neKOwV6Xd70wbLfhw1PQb8Z0OvHzRuoMabNsUTQ2pXvp2DzXEZsfYVVQ1YSvqWi8f2HPgFD/7fh7UU74D83QIcRkPp888ZqjGmTLBG0Vtnvw+Y/we7FdKyu4JTIRMqSbye8/1ToOBKoUzSk1ZD+U1j7CGglDH3s6GGgq8ph6VWgVXDmXAiNbKmzMca0YpYIWqPiLPhiMrRPIqfrrdzxxUmcM/ZH3Dd2YOOfG/MKEALrnnAu9sOePDIZrPo55C+Ds96F2JN9egrGmLbDEkFrtPU10Gqqzl3Cza/mcLBdBbeN9+LCLSEw5q8gobD+104yGP5/TjLInAubn4eB9zkTwBtjjMsSQWuj6oz332UCb28MZ2PuQV685lTaR3jZzl9CYPRfnF7CG55yion6zYBvboZOp8OI3/o0fGNM22OJoLXJWwqFGRQPeIg/zP2e0ckd+dHQ7sd2DAlxh4QOgY1/gIy/Qkg4nPm2M4y0McZ4sETQ2mz9O4TF8tzmYewr3s0jlw46vj4DIpD6gvNksPlFGP8eRPdu/niNMW2eT3sSicgkEfleRDJE5MF6tvcWkTQRWSUia0TkYl/G0+pVFELmXAq7TeGVb/Yw9bReDOkZf/zHE4FRz8KUfGciGWOMqYfPEoGIhAIzgYuAQcA0ERlUZ7f/Ad5R1VOBqcCffBVPm5A5FyqLeG3XBEJChHvPG9A8x404gWRijAl4vnwiGA1kqOpWVS0H5gCT6+yjQJz7Ph7I8WE8rd/Wv1MR1Z+nv+vMNaN70y3e2vkbY3zPl4mgJ7DTYznLXefpMeA6EckCFgF313cgEblVRNJFJD0vL88XsfrfwS2Qt5SPSi8mLCSEn07o5++IjDFBwt+jjU0DZqlqEnAx8IbI0SOgqerLqpqqqqmJiYktHmSL2DYLJYRfrR/J9af3oUucPQ0YY1qGLxNBNtDLYznJXefpZuAdAFX9DxAJdPZhTK1TdRVsfY0NjOUAXbhtvD0NGGNaji8TwXKgv4j0FZEInMrghXX2yQTOAxCRFJxEEHhlP6pQVtDw9l2fQEk2f9pxJjec0YfE2HYtF5sxJuj5LBGoaiVwF/ARsBGnddB6EXlCRC5zd7sfmCEiq4HZwHRVVV/F5DdbZ8G7nWHFfVBZVM/2VykkgaWlZ3Dr2Se1eHjGmODm0w5lqroIpxLYc90jHu83AON8GUOrsG0WhMXA989B9nsw5m/Q9RxnW1kB1TsXMHfvRK49YwCdYuxpwBjTsvxdWRz4inNgz5eQ8gCc/zkg8Om5sOwOqDgE298iRMt5v3ASM86ypwFjTMuzISZ8bec8QKH3VRB/Cly8xpkmctMzkLOIcg1lc0k/xqaeR4doGwfIGNPy7InA1zLfgYRhThIACIuCkX+EC76CsCgiSrbx3sELueVMexowxviHJQJfKtoJeV85TwN1JZ7B+uGfcev2/yZ6yJ3ER4W3fHzGGIMlAt/aOc/5WV8iAJ5ZspNvys9i+lnNNKaQMcYcB0sEvrTjbehwKsT1P2rTmqz9LN64mxlnnURcpD0NGGP8xxKBrxRuh/xvoc/V9W5+5pPNJESFM31ccouGZYwxdVki8JXMuc7P3j85atOqzH2kfZ/HrWefRKw9DRhj/MwSga9kvgMdUyHm6NZAzyzeQsfoCG48I7nl4zLGmDosEfjCoR+gIL3eYqEVOwr4YnMet519EtHtrBuHMcb/LBH4QiPFQs98soXOMRFcf0afFg7KGGPqZ4nAFzLfhk6nQ/SRF/tvt+azNGMvt4/vR1SEPQ0YY1oHSwTN7eBm2Pcd9Dm678AzizeTGNuOa8fY04AxpvWwRNDcMt9xftYpFvr6h718s7WAn07oR/uIUD8EZowx9WsyEYjIpfVNH2kakPkOJI6DqKTaVarKs59soWtcO6aN7u3H4Iwx5mjeXOCvBraIyO9E5BRfB9SmHdgI+9dC7yNbC32Vkc+y7QXcec7JRIbb04AxpnVpssZSVa8TkTjcieZFRIFXgdmqesjXAbZKax6F3I+OXl+WBwj0mnLE6ueXbKFbXCRXn9br6M8YY4yfeVXko6oHgXnAHKA7cDmwUkTu9mFsrVP5Plj/f1BxECISjnzF9ochj0BUj9rdv92az7JtBdw2/iTahdnTgDGm9WnyicCdX/gm4GTgdWC0qu4RkShgA/CCb0NsZXbOB62EsW9Ax1FN7v5iWgadYyKsbsAY02p505h9CvCMqn7huVJVi0XkZt+E1YplvuMMG9FhZJO7rsrcx5db9vLQRadY3YAxptXypmjoMWBZzYKItBeRZABV/dQnUbVWpXth12KnMlikyd1fXJJBQlQ4151u/QaMMa2XN4lgLlDtsVzlrgs+WfNBq+rtLFbXuuwDfLppDzeP62tjChljWjVvEkGYqpbXLLjvg3OW9R1vQ+wASBje5K4z0zKIjQzjRptvwBjTynmTCPLcCmMARGQysNd3IbVSJbthT5oz7WQTxUKbdx/ig3W7uGlsss0+Zoxp9bxJBLcDD4tIpojsBH4J3ObNwUVkkoh8LyIZIvJgPdufEZHv3NdmEdl/bOG3oKx/glY3OOOYp5lpGURHhHLTuL4tEJgxxpwYbzqU/QCcLiIx7nKhNwcWkVBgJnABkAUsF5GFqrrB49g/89j/buDUYwu/Be14G+JSIH5wo7tt21vEe6tzmHH2SXSIDs4SNGNM2+JVLaaI/AgYDESKWyyiqk808bHRQIaqbnWPMQeYjNP3oD7TgEe9iafFleTCni+czmJNFAv9KS2DiLAQbjnz6JnJjDGmNfJm0Lm/4Iw3dDcgwE8Ab9pD9gR2eixnuevq+44+QF9gSQPbbxWRdBFJz8vL8+Krm1nmPECbbC20s6CY+auymTa6N4mx7VomNmOMOUHe1BGMVdUbgH2q+jhwBjCgmeOYCsxT1ar6Nqrqy6qaqqqpiYmJzfzVXsh8B+KHQPygBndRVZ76YBMhItx2dr8WDM4YY06MN4mg1P1ZLCI9gAqc8Yaakg14jrKW5K6rz1RgthfHbHnFWZC3tMlK4pe+2Mq/1+Zy3wX96RYf2ULBGWPMifMmEbwnIgnA74GVwHbgLS8+txzoLyJ9RSQC52K/sO5O7tDWHYD/eBt0i8qc5/zs3XCxUNqmPfz2w01cMhpCfKAAAB14SURBVKw7d4y3pwFjTNvSaGWxOyHNp6q6H3hXRN4HIlX1QFMHVtVKEbkL+AgIBf6uqutF5AkgXVVrksJUYI6q6gmdia/seBs6jIC4+kvDtuYVcs+cVaR0i+P3Vw5HvBh6whhjWpNGE4GqVovITNxmnapaBpR5e3BVXQQsqrPukTrLj3l7vBZXtAPyv4Hh/1fv5oOlFcx4PZ3w0BBevmGUTUFpjGmTvCka+lREpkgw3upmukMq1VMsVFWt3DfnO3bkF/Ona0eS1CGqhYMzxpjm4U0iuA1nkLkyETkoIodE5KCP42oddrzjzDkQe3S5/9OffM+STXt45NJBnH5SJz8EZ4wxzaPJRKCqsaoaoqoRqhrnLse1RHB+VXEQCpZDz8lHbfr3mlxmpv3A1NN6cb0NMW2MaeO8maHs7PrW152oJuAUuX3hYvsften3H21iSM84Hp882CqHjTFtnjdDTPzc430kztARK4BzfRJRa1HsJoLoIyecP1Bcwfb8Yn4+caDNQWyMCQjeDDp3qeeyiPQCnvVZRK1FTSKIOjIRrM91Ws4O6Rnf0hEZY4xPeFNZXFcWkNLcgbQ6xTtBQqB9jyNWr8t2E0GPwK8mMcYEB2/qCF4Aajp7hQAjcHoYB7binRDZHUKO/BWtyz5Ij/hIOsXYoHLGmMDgTR1Busf7SmC2qn7lo3haj+IsiEo6avW6nAMMtmIhY0wA8SYRzANKa0YGFZFQEYlS1WLfhuZnxTshfugRqwrLKtm2t4jJw+sdTdsYY9okr3oWA+09ltsDi30TTiuh6jQfrVNRvCHnIKowNMnqB4wxgcObRBDpOT2l+z6wx1Mo3wdVxUc1HT1cUWxFQ8aYwOFNIigSkZE1CyIyCijxXUitQANNR9flHCAxth1d4my+AWNM4PCmjuA+YK6I5OBMVdkNZ+rKwNVQH4Lsg9Zs1BgTcLzpULbcnTxmoLvqe1Wt8G1YflabCA63Giopr2LLnkNcOLirn4Iyxhjf8Gby+juBaFVdp6rrgBgR+anvQ/Oj4iyQMIjsVrtq466DVKv1KDbGBB5v6ghmuDOUAaCq+4AZvgupFSja6fQoDjk8ltD6bBtawhgTmLxJBKGek9KISCgQ4buQWoHinfW0GDpIh6hwetjE9MaYAONNIvgQeFtEzhOR84DZwAe+DcvPio/uQ7Au5wBDesbbsNPGmIDjTSL4JbAEuN19reXIDmaBRdUdXuJwIiirrGLz7kNWLGSMCUjezFBWDXwLbMeZi+BcYKNvw/KjsjyoLjuixdDmXYVUVKl1JDPGBKQGm4+KyABgmvvaC7wNoKrntExoflKc5fz0eCJYl1NTUWx9CIwxgaexfgSbgC+BS1Q1A0BEftYiUflTPZ3J1mUfIDYyjN4dA3tkDWNMcGqsaOgKIBdIE5G/uhXFgV9TWlRPIsg5yJAeVlFsjAlMDSYCVV2gqlOBU4A0nKEmuojIn0XkQm8OLiKTROR7EckQkQcb2OcqEdkgIutF5K3jOYlmVbwTQiIgMhGAiqpqNuYetGIhY0zA8qayuEhV33LnLk4CVuG0JGqU299gJnARMAiYJiKD6uzTH3gIGKeqg3GSjX8V73QqisX51WTsKaS8stpaDBljAtYxzVmsqvtU9WVVPc+L3UcDGaq6VVXLgTnA5Dr7zABmur2VUdU9xxKPT9QkAlfN0NODrcWQMSZAHc/k9d7qCez0WM5y13kaAAwQka9E5BsRmVTfgUTkVhFJF5H0vLw8H4XrqtOHYH3OQaIiQunbOdq332uMMX7iy0TgjTCgPzABp5nqX0Ukoe5O7lNIqqqmJiYm+i4arYaS7CMSwdrsAwzuEUdoiFUUG2MCky8TQTbgOU5DkrvOUxawUFUrVHUbsBknMfhH6W6orqhNBFXVyoacg1YsZIwJaL5MBMuB/iLSV0QigKnAwjr7LMB5GkBEOuMUFW31YUyNq9N0dNveQkoqqqyi2BgT0HyWCFS1ErgL+AhnSIp3VHW9iDwhIpe5u30E5IvIBpwmqj9X1XxfxdSkms5k7sij67IPAtaj2BgT2LyZqvK4qeoiYFGddY94vFfg/7kv/6tJBO2dVkPrsg/QLiyEkxNj/BiUMcb4lr8ri1uX4iwIjYR2nQCnojilexxhofZrMsYELrvCeaqZh0CEarei2IqFjDGBzhKBJ48JafYWlXGorJL+XWL9HJQxxviWJQJPHokgZ38pAD0SAncOHmOMAUsEh1VXQklObSLI3V8CQI8Em6PYGBPYLBHUKMl1eha74wxl1ySCeHsiMMYENksENepMSJN7oJT24aEkRIX7MShjjPE9SwQ1aqaojK5JBCV0T4i0yWiMMQHPEkGNOk8E2ftLrVjIGBMULBHUKN4JYTEQ7owrlLu/xCqKjTFBwRJBDY/OZOWV1eQVltHdngiMMUHAEkGNosMzk+0+WIqqNR01xgQHSwQ1juhMVtOHwJ4IjDGBzxIBQFW5MymNR9NRwIqGjDFBwRIBOD2K0dqmo9nWq9gYE0QsEUA9nclKSIgKJyrCp9M1GGNMq2CJAI5KBDn7S61YyBgTNCwRgEcicFoN5ewvoacVCxljgoQlAnCajobHQ7gz90DuAXsiMMYED0sEACVZtcVCRWWVHCipoLs9ERhjgoQlAnA7kx2uKAboaX0IjDFBwhIBOHUE0UfOTGZFQ8aYYGGJoKoUyvKg/eGKYoDu8VY0ZIwJDj5NBCIySUS+F5EMEXmwnu3TRSRPRL5zX7f4Mp561ZmHIOdAKSLQzRKBMSZI+KzHlIiEAjOBC4AsYLmILFTVDXV2fVtV7/JVHE0q2u78jOoNOMNPd4ltR3ioPSwZY4KDL692o4EMVd2qquXAHGCyD7/v+BSsdH52GA5AzoESqx8wxgQVXyaCnsBOj+Usd11dU0RkjYjME5FePoynfgXpEN0X2nUCIHd/qbUYMsYEFX+Xf7wHJKvqMOAT4LX6dhKRW0UkXUTS8/LymjeC/HTolAqAqrpPBFY/YIwJHr5MBNmA5x1+kruulqrmq2qZu/g3YFR9B1LVl1U1VVVTExMTmy/Csnwo2gYdnUSwr7iC0opqm4fAGBNUfJkIlgP9RaSviEQAU4GFnjuISHePxcuAjT6M52gFK5yfbiLIseGnjTFByGethlS1UkTuAj4CQoG/q+p6EXkCSFfVhcA9InIZUAkUANN9FU+9CtKdnx1HAp59COyJwBgTPHw64L6qLgIW1Vn3iMf7h4CHfBlDo/LTIbY/RCQAh2cms6IhY0ww8XdlsX8VrICOh6slcg6UEBEaQqfoCD8GZYwxLSt4E0HpHijOrK0fAGecoW7xkYSEiB8DM8aYlhW8iaBORTE4vYqtotgYE2yCNxHkpwMCHU+tXZV7oJQeVlFsjAkywZsICtIhbiCExwFQVa3sOlhqE9IYY4JOcCcCj2KhPYdKqapWazFkjAk6wZkISnKhJOeoimLAioaMMUHHp/0IWq2aiuJOnomgplexJQLTelRUVJCVlUVpaam/QzFtRGRkJElJSYSHh3v9meBMBPnpICHQYUTtqpq5iq2OwLQmWVlZxMbGkpycjIg1azaNU1Xy8/PJysqib9++Xn8uOIuGCtIhLgXComtX5ewvJaZdGHGR3mdRY3yttLSUTp06WRIwXhEROnXqdMxPkMGXCFSPqigGp2jI+hCY1siSgDkWx/P3EnyJoCQbSncflQhyD5TaYHPGmKAUfIkg3x1xtJM9ERjjrQULFiAibNq0yd+hHJfc3FwuueQSPvroI0aMGMGIESOIiYlh4MCBjBgxghtuuMGr4/zlL3/h9ddfb3SftWvXMn369GaIuuUEX2VxQTpIKCQMr11VWlFFflG5NR01pgGzZ8/mzDPPZPbs2Tz++OM++56qqipCQ0Ob/bhPP/00M2bMYOLEiUycOBGACRMm8Ic//IHU1CNvChuL4fbbb2/yu4YOHUpWVhaZmZn07t37xINvAcGZCOKHQNjhi/4ud/jp7tZ01LRij7+3ng05B5v1mIN6xPHopYMb3aewsJClS5eSlpbGpZdeWpsIqqqq+OUvf8mHH35ISEgIM2bM4O6772b58uXce++9FBUV0a5dOz799FPeffdd0tPTefHFFwG45JJLeOCBB5gwYQIxMTHcdtttLF68mJkzZ7JkyRLee+89SkpKGDt2LC+99BIiQkZGBrfffjt5eXmEhoYyd+5cHn/8ca644gp+/OMfA3Dttddy1VVXMXny5CPO4d133+VXv/pVg+eYnJzM1VdfzSeffMIvfvELDh06xMsvv0x5eTknn3wyb7zxBlFRUTz22GPExMTUxj5mzBjS0tLYv38/r7zyCmeddRYAl156KXPmzOEXv/jFcf/btKTgKhpSdfoQ1FMsBNDD5io25ij/+te/mDRpEgMGDKBTp06sWOH0w3n55ZfZvn073333HWvWrOHaa6+lvLycq6++mueee47Vq1ezePFi2rdv/AarqKiIMWPGsHr1as4880zuuusuli9fzrp16ygpKeH9998HnIv8nXfeyerVq/n666/p3r07N998M7NmzQLgwIEDfP311/zoRz864vjbtm2jQ4cOtGvXrtE4OnXqxMqVK5k6dSpXXHEFy5cvZ/Xq1aSkpPDKK6/U+5nKykqWLVvGs88+e8STUmpqKl9++WWj39eaBNcTQXEmlO09usWQTUhj2oCm7tx9Zfbs2dx7770ATJ06ldmzZzNq1CgWL17M7bffTliYcxnp2LEja9eupXv37px22mkAxMXFNXn80NBQpkyZUruclpbG7373O4qLiykoKGDw4MFMmDCB7OxsLr/8csDpNAUwfvx4fvrTn5KXl8e7777LlClTauOpkZubizdznV999dW179etW8f//M//sH//fgoLC2uLk+q64oorABg1ahTbt2+vXd+lSxdycnKa/M7WIrgSQU1Fcd0WQ+4TQTd7IjDmCAUFBSxZsoS1a9ciIlRVVSEi/P73vz+m44SFhVFdXV277NnOPTIysrZMvrS0lJ/+9Kekp6fTq1cvHnvssSbbxN9www384x//YM6cObz66qtHbW/fvr1X7eqjow/3K5o+fToLFixg+PDhzJo1i88++6zez9Q8ZYSGhlJZWXnE+TX1JNSaBFfRUEE6hIRDwtAjVuccKKFzTASR4c1fSWVMWzZv3jyuv/56duzYwfbt29m5cyd9+/blyy+/5IILLuCll16qvQAWFBQwcOBAcnNzWb58OQCHDh2isrKS5ORkvvvuO6qrq9m5cyfLli2r9/tqLtidO3emsLCQefPmARAbG0tSUhILFiwAoKysjOLiYsC5aD/77LMADBo06KhjDhgw4Ii7dW8cOnSI7t27U1FRwZtvvnlMnwXYvHkzQ4YMOebP+UvwJYL4oRB6ZFlhzn7rQ2BMfWbPnl1bHFNjypQpzJ49m1tuuYXevXszbNgwhg8fzltvvUVERARvv/02d999N8OHD+eCCy6gtLSUcePG0bdvXwYNGsQ999zDyJEj6/2+hIQEZsyYwZAhQ5g4cWJtERPAG2+8wfPPP8+wYcMYO3Ysu3btAqBr166kpKRw00031XvM6Oho+vXrR0ZGhtfn/eSTTzJmzBjGjRvHKaec4vXnaqSlpR1VV9Gaiar6O4Zjkpqaqunp6cf+QVWY1xH6XAWjXzpi0wVPf07fztG8fENqAx82xj82btxISkqKv8No1YqLixk6dCgrV64kPj6+3n3mz5/PihUrGm051FzKysoYP348S5cuPaq+oqXU93cjIitUtd6LXPA8ERRuhYr9R9UPgDszmVUUG9PmLF68mJSUFO6+++4GkwDA5ZdfTnJycovElJmZyVNPPeW3JHA82k6kJ6qg/orig6UVFJZVWq9iY9qg888/nx07dni17y233OLjaBz9+/enf//+LfJdzSV4nghK8yCiA8Qf2QSvpg+B1REYY4KVTxOBiEwSke9FJENEHmxkvykioiLiu0L6gXfBlL0QGnHE6n+uzEbE6WFpjDHByGeJQERCgZnARcAgYJqIHNW2S0RigXuBb30Vy+EvO/J0t+YV8upX27hqVC/6Jcb4/OuNMaY18uUTwWggQ1W3qmo5MAeYXM9+TwK/BVp8Lr5f/Xsj7cJCeWDiwJb+amOMaTV8mQh6Ajs9lrPcdbVEZCTQS1X/3diBRORWEUkXkfS8vLxmCS7t+z0s2bSHe847mcTYxscgMSbYBcow1MXFxXTq1ImDB48cvO/HP/4xb7/9doOfj4lxSgxycnK48sor691nwoQJNNW0/dlnn63tCAdw8cUXs3//fm9P4wgPPPAAS5YsOa7P1uW3ymIRCQGeBu5val9VfVlVU1U11ZsxQ5pSUVXNk+9voG/naKaP9X5eT2OClecw1L5UVVXlk+PWDEMdFRXFxIkTmT9/fu22AwcOsHTpUi699NImj9OjR4/a3s7Ho24iWLRoEQkJCcd1rLvvvpunnnrquGPx5MtEkA308lhOctfViAWGAJ+JyHbgdGChTyuMXa//Zwdb84r430tSiAgLnoZTpo1bcR8sntC8rxX3Nfm1NcNQv/LKK8yZM6d2fVVVFQ888ABDhgxh2LBhvPDCCwAsX76csWPHMnz4cEaPHs2hQ4eYNWsWd911V+1nL7nkktrxe2JiYrj//vsZPnw4//nPf3jiiSc47bTTGDJkCLfeeis1nV4zMjI4//zzGT58OCNHjuSHH37ghhtuqB12ApwRSv/1r38ddQ7vvvsukyZNAmDatGlHnMf8+fOZOHEi1dXVnHfeeYwcOZKhQ4fWe5zt27fXDh1RUlLC1KlTSUlJ4fLLL6ekpKR2vzvuuIPU1FQGDx7Mo48+CsDzzz9PTk4O55xzDueccw7gDH+9d+9ewElWQ4YMYciQIbVDZmzfvp2UlBRmzJjB4MGDufDCC2u/p0+fPuTn59f2sD4RvrwKLgf6i0hfEYkApgILazaq6gFV7ayqyaqaDHwDXKaqx9Ft2Hv5hWU8u3gz4wckcs7ALr78KmMCQqANQz1x4kRWrlxJfn4+AHPmzGHatGlERkYyf/58Vq5cSVpaGvfffz+Njbzw5z//maioKDZu3Mjjjz9e+3sB+PWvf016ejpr1qzh888/Z82aNdxzzz306NGDtLQ00tLSjjjWihUrePXVV/n222/55ptv+Otf/8qqVasA2LJlC3feeSfr168nISGBd999t/ZzI0eO5Kuvvmr09+sNn3UoU9VKEbkL+AgIBf6uqutF5AkgXVUXNn4E3/jDx5spKa/ify9JsUnBTdsy6lm/fG2gDUMdERHBZZddxrx585gyZQqrVq1i4sSJqCoPP/wwX3zxBSEhIWRnZ7N79266detWb9xffPEF99xzDwDDhg1j2LBhtdveeecdXn75ZSorK8nNzWXDhg1HbK9r6dKlXH755bUjoF5xxRV8+eWXXHbZZfTt25cRI0YAvhvu2qc9i1V1EbCozrpHGth3gi9jAVifc4A5yzO5aWxfTu4S6+uvM6bNC9RhqKdNm8aTTz6JqjJ58mTCw8OZNWsWeXl5rFixgvDwcJKTk70avrqubdu28Yc//IHly5fToUMHpk+fflzHqeE5oU5oaOgRRVDNNdx10BSQqyqPv7eBDlER3Ht+2+r+bYy/BOow1BMmTGDLli3MnDmTadOmAU7RUpcuXQgPDyctLa3JoSvOPvts3nrrLcCZyGbNmjUAHDx4kOjoaOLj49m9ezcffPBB7WdiY2M5dOjQUcc666yzWLBgAcXFxRQVFTF//vzaaS8b01zDXQdNIli0dhfLthXwwIUDiW8f7u9wjGkTAnUY6pCQEK688kry8/MZP3484NRBpKenM3ToUF5//fUmh5++4447KCwsJCUlhUceeYRRo0YBMHz4cE499VROOeUUrrnmGsaNG1f7mVtvvZVJkybVVhbXGDlyJNOnT2f06NGMGTOGW265hVNPPbXR76+oqCAjI4PU1BNvXxM0w1Cnfb+Ht77N5C/XjSI0xOoGTNtgw1A3rbUNQ91Saiq2n3zyyaO22TDUDThnYBf+ekOqJQFjAkhrHIa6pVRWVnL//U12w/JK8AxDbYwJOK1xGOqW8pOf/KTZjhU0TwTGtFVtrfjW+Nfx/L1YIjCmFYuMjCQ/P9+SgfGKqpKfn1/bz8JbVjRkTCuWlJREVlYWzTXYogl8kZGRJCUlHdNnLBEY04qFh4fTt68NjGh8y4qGjDEmyFkiMMaYIGeJwBhjglyb61ksInlAUw2HOwN7WyCc1sbOO7gE63lD8J77iZx3H1Wtd2avNpcIvCEi6Q11pQ5kdt7BJVjPG4L33H113lY0ZIwxQc4SgTHGBLlATQQv+zsAP7HzDi7Bet4QvOfuk/MOyDoCY4wx3gvUJwJjjDFeskRgjDFBLuASgYhMEpHvRSRDRB70dzy+IiJ/F5E9IrLOY11HEflERLa4Pzv4M0ZfEJFeIpImIhtEZL2I3OuuD+hzF5FIEVkmIqvd837cXd9XRL51/97fFpEIf8fqCyISKiKrROR9dzngz1tEtovIWhH5TkTS3XU++TsPqEQgIqHATOAiYBAwTUSOns06MMwCJtVZ9yDwqar2Bz51lwNNJXC/qg4CTgfudP+NA/3cy4BzVXU4MAKYJCKnA78FnlHVk4F9wM1+jNGX7gU2eiwHy3mfo6ojPPoO+OTvPKASATAayFDVrapaDswBJvs5Jp9Q1S+AgjqrJwOvue9fA37cokG1AFXNVdWV7vtDOBeHngT4uauj0F0Md18KnAvMc9cH3HkDiEgS8CPgb+6yEATn3QCf/J0HWiLoCez0WM5y1wWLrqqa677fBXT1ZzC+JiLJwKnAtwTBubvFI98Be4BPgB+A/apa6e4SqH/vzwK/AKrd5U4Ex3kr8LGIrBCRW911Pvk7t/kIApSqqogEbNtgEYkB3gXuU9WDzk2iI1DPXVWrgBEikgDMB07xc0g+JyKXAHtUdYWITPB3PC3sTFXNFpEuwCcisslzY3P+nQfaE0E20MtjOcldFyx2i0h3APfnHj/H4xMiEo6TBN5U1X+6q4Pi3AFUdT+QBpwBJIhIzQ1dIP69jwMuE5HtOEW95wLPEfjnjapmuz/34CT+0fjo7zzQEsFyoL/boiACmAos9HNMLWkhcKP7/kbgX36MxSfc8uFXgI2q+rTHpoA+dxFJdJ8EEJH2wAU49SNpwJXubgF33qr6kKomqWoyzv/nJap6LQF+3iISLSKxNe+BC4F1+OjvPOB6FovIxThliqHA31X1134OySdEZDYwAWdY2t3Ao8AC4B2gN85Q3Vepat0K5TZNRM4EvgTWcrjM+GGceoKAPXcRGYZTORiKcwP3jqo+ISIn4dwpdwRWAdepapn/IvUdt2joAVW9JNDP2z2/+e5iGPCWqv5aRDrhg7/zgEsExhhjjk2gFQ0ZY4w5RpYIjDEmyFkiMMaYIGeJwBhjgpwlAmOMCXKWCIxxiUiVO9JjzavZBq4TkWTPkWKNaU1siAljDitR1RH+DsKYlmZPBMY0wR0X/nfu2PDLRORkd32yiCwRkTUi8qmI9HbXdxWR+e7cAatFZKx7qFAR+as7n8DHbg9hROQed36FNSIyx0+naYKYJQJjDmtfp2joao9tB1R1KPAiTs91gBeA11R1GPAm8Ly7/nngc3fugJHAend9f2Cmqg4G9gNT3PUPAqe6x7ndVydnTEOsZ7ExLhEpVNWYetZvx5kUZqs74N0uVe0kInuB7qpa4a7PVdXOIpIHJHkOeeAOmf2JO6EIIvJLIFxVfyUiHwKFOEOELPCYd8CYFmFPBMZ4Rxt4fyw8x8Kp4nAd3Y9wZtYbCSz3GFXTmBZhicAY71zt8fM/7vuvcUbEBLgWZzA8cKYQvANqJ5OJb+igIhIC9FLVNOCXQDxw1FOJMb5kdx7GHNbenQGsxoeqWtOEtIOIrMG5q5/mrrsbeFVEfg7kATe56+8FXhaRm3Hu/O8AcqlfKPAPN1kI8Lw734AxLcbqCIxpgltHkKqqe/0dizG+YEVDxhgT5OyJwBhjgpw9ERhjTJCzRGCMMUHOEoExxgQ5SwTGGBPkLBEYY0yQ+/8xANk8OLuGCwAAAABJRU5ErkJggg==\n" }, "metadata": { "needs_background": "light" } } ] }, { "cell_type": "code", "source": [ "print(y_test[10])" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "53gqb6ET2Aqt", "outputId": "0b2b313b-3dc2-45e8-9749-dae496f32b4f" }, "execution_count": 33, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "[0. 0. 1.]\n" ] } ] }, { "cell_type": "code", "source": [ "x_test_pattern = X_test[10,:]\n", "y_pred = multi_classifier.predict(x_test_pattern.reshape(1,-1))\n", "print(y_pred[0])" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "uGCfokXP2AvA", "outputId": "8a213786-aeae-4e59-a46e-85b5ab29634c" }, "execution_count": 34, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "1/1 [==============================] - 0s 53ms/step\n", "[1.13614085e-04 4.17691655e-02 9.58117247e-01]\n" ] } ] } ] }