Hands-On Machine Learning for Algorithmic Trading

How to further improve the results

The relatively simple architecture yields some promising results. To further improve performance, you can do the following:

• First and foremost, add new features and more data to the model
• Expand the set of architectures to explore, including more or wider layers
• Inspect the training progress and train for more epochs if the validation error continued to improve at 50 epochs

Finally, you can use more sophisticated architectures, including RNNs and CNNs, that are well suited to sequential data, whereas vanilla feedforward neural networks are not designed to capture the ordered nature of the features.

We will turn to these specialized architectures in the following chapter.