Hands-On Machine Learning for Algorithmic Trading

Alternative RNN architectures

Just like the feedforward and CNN architectures we covered in the previous two chapters, RNNs can be designed in a variety of ways that best capture the functional relationship and dynamic between input and output data.

In addition to the recurrent connections between the hidden states, there are several alternative approaches, including recurrent output relationships, bidirectional RNNs, and encoder-decoder architectures (you can refer to GitHub for more background references to complement this brief summary here: https://github.com/PacktPublishing/Hands-On-Machine-Learning-for-Algorithmic-Trading).