Machine Learning in Elixir

Wrapping Up

In this chapter, you were introduced to deep learning and why it’s much more effective than other approaches to certain classes of problems. You broke down the anatomy of a neural network and implemented one of the most common types of neural networks: the basic feed-forward or dense network. You used Axon to create, train, and evaluate neural networks.

While feed-forward networks can be used to solve a number of complex problems, there are certain classes of problems where feed-forward networks are not enough. In the next few chapters, you’ll solve problems in different areas where feed-forward networks come up short. You’ll also implement different types of neural networks that overcome the limitations of basic feed-forward networks ...