Initial commit

main.py

@@ -0,0 +1,124 @@
+import numpy as np
+import pandas as pd
+import tensorflow as tf
+import tensorflow_addons as tfa
+from sklearn.model_selection import train_test_split
+from sklearn.metrics import f1_score, recall_score, log_loss, precision_score
+from sklearn.preprocessing import StandardScaler
+from sklearn.naive_bayes import GaussianNB
+
+
+def main():
+    no_of_epochs = 50
+    batch_size = 64
+
+    bayes = GaussianNB()
+
+    feature_names = ["BMI", "SleepTime", "Sex", "Diabetic", "PhysicalActivity", "Smoking", "AlcoholDrinking",
+                     "HeartDisease", "KidneyDisease", "Stroke", "Asthma"]
+
+    # feature_names = ["Sex", "KidneyDisease"]
+    train_feature_names = ["BMI", "SleepTime", "Sex", "Diabetic", "PhysicalActivity", "Smoking", "AlcoholDrinking",
+                      "KidneyDisease", "Stroke", "Asthma"]
+
+    dataset = pd.read_csv('heart_2020_cleaned.csv')
+    dataset = dataset.dropna()
+
+    dataset = dataset[feature_names]
+
+    dataset["Diabetic"] = dataset["Diabetic"].apply(lambda x: int("Yes" in x))
+    dataset["HeartDisease"] = dataset["HeartDisease"].apply(lambda x: int(x == "Yes"))
+    dataset["KidneyDisease"] = dataset["KidneyDisease"].apply(lambda x: int(x == "Yes"))
+    dataset["PhysicalActivity"] = dataset["PhysicalActivity"].apply(lambda x: int(x == "Yes"))
+    dataset["Stroke"] = dataset["Stroke"].apply(lambda x: int(x == "Yes"))
+    dataset["Smoking"] = dataset["Smoking"].apply(lambda x: int(x == "Yes"))
+    dataset["Asthma"] = dataset["Asthma"].apply(lambda x: int(x == "Yes"))
+    dataset["AlcoholDrinking"] = dataset["AlcoholDrinking"].apply(lambda x: int(x == "Yes"))
+    dataset["Sex"] = dataset["Sex"].apply(lambda x: 1 if x == "Female" else 0)
+
+    #PRZYGOTOWANIE DANYCH Z ODPOWIEDNIM PODZIAŁEM, PONIEWAŻ W ORYGINALNYCH DANYCH KLASA 0 STANOIWŁA 91% PRZYKŁADÓW
+    classes_counts = dataset["HeartDisease"].value_counts()
+    positive_counts = classes_counts.get(1)
+    offset = int(round(positive_counts * 3, 0))
+    dataset = dataset.sort_values(by=["HeartDisease"], ascending=False)[:offset]
+    dataset = dataset.sample(frac=1)
+
+    dataset_train, dataset_test = train_test_split(dataset, test_size=.3, train_size=.7, random_state=42)
+
+
+    scaler = StandardScaler()
+
+    model = tf.keras.Sequential([
+        tf.keras.layers.Dense(64, activation='relu'),
+        tf.keras.layers.Dense(32, activation='relu'),
+        tf.keras.layers.Dense(4, activation='relu'),
+        tf.keras.layers.Dense(1, activation='sigmoid')
+    ])
+
+    model2 = tf.keras.Sequential([
+        tf.keras.layers.Dense(4, activation='relu'),
+        tf.keras.layers.Dense(1, activation='sigmoid')
+    ])
+
+    model.compile(
+        loss=tf.keras.losses.binary_crossentropy,
+        optimizer='sgd',
+        metrics=[tf.keras.metrics.Precision(name="precision"),
+                 "accuracy",
+                 tf.keras.metrics.Recall(name='recall')]
+    )
+
+    model2.compile(
+        loss=tf.keras.losses.msle,
+        optimizer='sgd',
+        metrics=[tf.keras.metrics.Precision(name="precision"),
+                 "accuracy",
+                 tf.keras.metrics.Recall(name='recall')]
+    )
+
+    train_X = dataset_train[train_feature_names].astype(np.float32)
+    train_Y = dataset_train["HeartDisease"]
+
+    test_X = dataset_test[train_feature_names].astype(np.float32)
+    test_Y = dataset_test["HeartDisease"]
+
+    train_X = scaler.fit_transform(train_X)
+
+    test_X = scaler.fit_transform(test_X)
+
+    train_X = tf.convert_to_tensor(train_X)
+    # train_Y = tf.convert_to_tensor(train_Y)
+
+    test_X = tf.convert_to_tensor(test_X)
+    # test_Y = tf.convert_to_tensor(test_Y)
+
+    print("================ MODEL 1 TRAINING =======================")
+    model.fit(train_X, train_Y, epochs=no_of_epochs, batch_size=batch_size)
+
+    print("================ MODEL 2 TRAINING =======================")
+    model2.fit(train_X, train_Y, epochs=no_of_epochs, batch_size=batch_size)
+
+
+    prediction_1 = np.round(model.predict(test_X), 0)
+    prediction_2 = np.round(model2.predict(test_X), 0)
+
+    evaluation = model.evaluate(test_X, test_Y, batch_size=batch_size, return_dict=True)
+    evaluation_2 = model2.evaluate(test_X, test_Y, batch_size=batch_size, return_dict=True)
+
+    f1_model_1 = f1_score(prediction_1, test_Y)
+    f1_model_2 = f1_score(prediction_2, test_Y)
+
+    print(f"MODEL 1 EVALUATION: LOSS:{round(evaluation.get('loss'), 4)}, ACCURACY: {round(evaluation.get('accuracy'), 4)}, RECALL: {round(evaluation.get('recall'), 4)}, F1_SCORE:{round(f1_model_1, 4)}, PRECISION: {round(evaluation.get('precision'), 4)}")
+    print(f"MODEL 2 EVALUATION: LOSS:{round(evaluation_2.get('loss'), 4)}, ACCURACY: {round(evaluation_2.get('accuracy'), 4)}, RECALL: {round(evaluation_2.get('recall'), 4)}, F1_SCORE:{round(f1_model_2, 4)}, PRECISION: {round(evaluation_2.get('precision'), 4)}")
+    bayes.fit(train_X, train_Y)
+
+    pred_bayes = bayes.predict(test_X)
+    score_bayes = bayes.score(test_X, test_Y)
+    loss_bayes = log_loss(pred_bayes, test_Y)
+    precision_bayes = precision_score(pred_bayes, test_Y)
+    bayes_recall = recall_score(pred_bayes, test_Y)
+    bayes_f1 = f1_score(pred_bayes, test_Y)
+
+    print(f"NAIVE BAYES CLASSIFIER: LOSS: {loss_bayes} ,ACCURACY:{score_bayes}, RECALL: {bayes_recall},F1_SCORE: {bayes_f1}, PRECISION: {precision_bayes}")
+
+main()