Fig. 1.

LSTM unit structure. The input is fed into LSTM units with 64 hidden units followed by a final dense layer. The input gate decides which values will be updated and creates a vector of new values to be added and updated to the state. After data input, the LSTM’s forget gate decides which information to discard. This gate examines the prior hidden state ($h$) and current input, yielding a binary output. Subsequently, the LSTM decides what new information to store in the cell state. Finally, the LSTM unit decides sequential output, which is based on the current cell state. The sigmoid and hyperbolic activation functions determine which parts of the cell state to output.