Keras Deep Learning Cookbook

LSTM networks

LSTM is designed to avoid the long-term dependency problem. It remembers the information for a longer period of time.

All recurrent neural networks have the form of a chain of repeating modules of a neural network. In standard RNNs, this repeating module will have a very simple structure, such as a single tanh layer. LSTMs also have this chain-like structure, but the repeating module has a different structure.

There are four layers, interacting in a very special way, as shown in the following diagram:

Reference: http://colah.github.io/posts/2015-08-Understanding-LSTMs/

We will not go into more detail on how an LSTM works, but ...