def get_model(batch_size, window_size): model = Sequential() model.add( LSTM(64, batch_input_shape=(batch_size, window_size, 3), return_sequences=False, stateful=False)) model.add(Activation('relu')) model.Add(Dropout(0.25)) model.add(Dense(6)) model.add(Activation('softmax')) model.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy']) return model
filename_mlp = 'allFeatures.csv' dataframe_mlp = read_csv(filename_mlp) array_mlp = dataframe_mlp.values Y_mlp = array_mlp[:, 0:71] X_train, X_test, Y_train, Y_test = train_test_split(X, Y_mlp, stratify=Y_mlp, test_size=0.5, random_state=0) X_train_extracted = X_train[:, indices[0:1000]] X_test_extracted = X_test[:, indices[0:1000]] model = Sequential() model.Add( Dense(500, kernel_initializer='normal', activation='relu', input_dim=1000)) model.add(Dropout(0.5)) model.Add(Dense(250, kernel_initializer='normal', activation='relu')) model.add(Dropout(0.5)) model.Add(Dense(125, kernel_initializer='normal', activation='relu')) model.add(Dropout(0.5)) model.Add(Dense(71, kernel_initializer='normal', activation='sigmoid')) model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy']) model.fit(X_train_extracted, Y_train, epochs=50, batch_size=100) scores = model.evaluate(X_test_extracted, Y_test)