An LSTM neural network for time series

NN_LSTM_TS

# Deep learning packages
from keras.models import Sequential
from keras.layers import LSTM, Dense

# Defining the number of neurons in the LSTM layer
n_layer = 50

# Defining how many lags will be used in the time series
n_lag = 3

# Defining the model
model = Sequential()
model.add(LSTM(n_layer, activation='relu', input_shape=(n_lag, 1)))
model.add(Dense(1))

# The objective functions which will be minimized is mean squared error (mse)
model.compile(optimizer='adam', loss='mse')