In this chapter, we explored DBNs and looked at how these could be used to build classification pipelines using one or more RBM layers. We looked at various parameters within the RBM layer and their effects on accuracy, reconstruction loss, and test set accuracy. We also looked at single layer and multilayer DBNs using one or more RBMs.

In the next chapter we look at Generative models and how they differ from discriminative models.

Get Neural Network Programming with TensorFlow now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.