2-3-deep-bidirectional-rnn.py

import numpy as np

from sklearn.metrics import accuracy_score
from keras.datasets import reuters
from keras.preprocessing.sequence import pad_sequences
from keras.utils import to_categorical
from keras.models import Sequential
from keras.layers import Dense, LSTM, Activation, Bidirectional
from keras import optimizers
from keras.wrappers.scikit_learn import KerasClassifier

# parameters for data load
num_words = 30000
maxlen = 50
test_split = 0.3

(X_train, y_train), (X_test, y_test) = reuters.load_data(num_words = num_words, maxlen = maxlen, test_split = test_split)

# pad the sequences with zeros 
# padding parameter is set to 'post' => 0's are appended to end of sequences
X_train = pad_sequences(X_train, padding = 'post')
X_test = pad_sequences(X_test, padding = 'post')

X_train = np.array(X_train).reshape((X_train.shape[0], X_train.shape[1], 1))
X_test = np.array(X_test).reshape((X_test.shape[0], X_test.shape[1], 1))

y_data = np.concatenate((y_train, y_test))
y_data = to_categorical(y_data)
y_train = y_data[:1395]
y_test = y_data[1395:]

def deep_bidirectional_lstm():
    model = Sequential()
    model.add(Bidirectional(LSTM(10, return_sequences = True), input_shape = (49,1)))
    model.add(Bidirectional(LSTM(10, return_sequences = True)))
    model.add(Bidirectional(LSTM(10, return_sequences = True)))
    model.add(Bidirectional(LSTM(10, return_sequences = False)))
    model.add(Dense(46))
    model.add(Activation('softmax'))
    
    adam = optimizers.Adam(lr = 0.001)
    model.compile(loss = 'categorical_crossentropy', optimizer = adam, metrics = ['accuracy'])
    
    return model

model = KerasClassifier(build_fn = deep_bidirectional_lstm, epochs = 200, batch_size = 50, verbose = 1)
model.fit(X_train, y_train)

y_pred = model.predict(X_test)
y_test_ = np.argmax(y_test, axis = 1)

print(accuracy_score(y_pred, y_test_))