Generative Modelling
Table of Contents
1. Abstract
- Learning the underlying distribution of a dataset(domain).
- injecting randomness allows generating multiple instances from the same semantically close Latent Variable
- Using Supervised Learning in conjuction with Unsupervised Learning yields stronger results.
- see Semi-Supervised Learning
2. Relevant Nodes
2.1. GANs
2.2. Variational AutoEncoders
- see the notion of Latent Variables