In this chapter, we will finish our journey through the world of unsupervised learning, discussing some very popular neural models that can be employed to perform a data generating process and new samples that can be drawn from it. Moreover, we will analyze the functionality of self-organizing maps, which can adapt their structures so that specific units become responsive to distinct input patterns.
In particular, we will discuss the following topics:
- Generative adversarial networks (GANs)
- Deep convolutional GANs (DCGANs)
- Wasserstein GANs (WGANs)
- Self-organizing maps (SOMs)
