Sentimental Analysis

sentimental_lstm.py

# Training Module of LSTM 2 layer stack

max_review_length = 600
X_train = sequence.pad_sequences(X_train, maxlen=max_review_length)
X_test = sequence.pad_sequences(X_test, maxlen=max_review_length)

# To train the Model
def trainModel(model):

    history = model.fit(numpy.array(X_train), numpy.array(y_train), \
                validation_split=0.33, epochs=30, batch_size=64)

    # Final evaluation of the model
    scores = model.evaluate(numpy.array(X_test), numpy.array(y_test), verbose=0)
    print("Accuracy: %.2f%%" % (scores[1]*100))
    model.save('getEmotions_2LSTM.h5')

# Model Definition
embedding_vector_length = 32
model = Sequential()
model.add(Embedding(top_words_count, embedding_vector_length, input_length=max_review_length))
model.add(Dropout(0.75))
model.add(LSTM(200, bias_regularizer=L1L2(l1=0.0, l2=0.05), return_sequences=True))
model.add(Dropout(0.75))
model.add(LSTM(150, bias_regularizer=L1L2(l1=0.0, l2=0.05)))
model.add(Dropout(0.5))
model.add(Dense(1, activation='sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])

trainModel(model=model)