def train_plot(training_data):
#plot the training data
plt.plot(training_data['loss'], linewidth=2, label='Train')
plt.plot(training_data['val_loss'], linewidth=2, label='Valid')
plt.legend(loc='upper right')
plt.title('Model Loss')
plt.ylabel('Loss')
plt.xlabel('Epoch')
plt.show()
if __name__ == '__main__':
#import data
df = pd.read_csv('Book1.csv', index_col = 0)
x_train, x_valid, y_train, y_valid, X, y = preprocessing(df)
input_dim = x_train.shape[1]
#create an instance of the model class
model = DNN(input_dim)
model.summary()
#set save path and train
model.set_save_path('dnn_v1.h5')
output = model.fit(x_train, y_train, x_valid, y_valid)
train_plot(output)
#load model and predict
model.load_model('dnn_v1.h5')
result = model.predict(X)
else:
nozerotrain = nozerotrain + 1
train_label_new.append(1)
prefilter_train = np.concatenate((train1, train2), axis=1)
prefilter_test = np.concatenate((test1, test2), axis=1)
train_label_new_forDNN = np.array([[0, 1] if i == 1 else [1, 0] for i in train_label_new])
test_label_new_forDNN = np.array([[0, 1] if i == 1 else [1, 0] for i in real_labels])
model_DNN = DNN()
model_DNN.fit(prefilter_train, train_label_new_forDNN, epochs=30, batch_size=200)
model_DNN.save("my_model")
# model_DNN = models.load_model("my_model")
print(model_DNN.summary())
print(model_DNN.summary())
proba = model_DNN.predict_classes(prefilter_test, batch_size=200, verbose=True)
ae_y_pred_prob = model_DNN.predict_proba(prefilter_test, batch_size=200, verbose=True)
acc, precision, sensitivity, specificity, MCC = calculate_performace(len(real_labels), proba, real_labels)
fpr, tpr, auc_thresholds = roc_curve(real_labels, ae_y_pred_prob[:, 1])
auc_score = auc(fpr, tpr)
precision1, recall, pr_threshods = precision_recall_curve(real_labels, ae_y_pred_prob[:, 1])
aupr_score = auc(recall, precision1)
# f = f1_score(real_labels, transfer_label_from_prob(ae_y_pred_prob[:,1]))
all_F_measure = np.zeros(len(pr_threshods))
print(auc_score)
print("+++++")
print(aupr_score)
