Machine Learning in Elixir

Wrapping Up

In this chapter, you implemented three types of deep unsupervised learning and deep generative modeling algorithms in Axon: an autoencoder, a variational autoencoder, and a generative adversarial network. You implemented custom event handlers in Axon and wrote your first custom training loop. You also used some new Axon layers such as Axon.conv_transpose/3 and Axon.resize/3.

Finally, you learned a bit about unsupervised learning and generative modeling in theory and practice. You also learned about the state of the art in generative modeling and about some applications of generative models in the real world.

At this point, you’ve learned essentially everything you need to know about deep learning and Axon. You’re more than ready ...