Evaluating the Disentanglement of Deep Generative Models through Manifold Topology.
Sharon ZhouEric ZelikmanFred LuAndrew Y. NgStefano ErmonPublished in: CoRR (2020)
Keyphrases
- generative model
- probabilistic model
- discriminative learning
- prior knowledge
- topic models
- deep belief networks
- conditional random fields
- em algorithm
- mixture model
- discriminative models
- hierarchical hidden markov models
- generative and discriminative models
- semi supervised
- maximum entropy principle
- object categories
- discriminative methods
- expectation maximization
- language model
- latent variables
- sparse coding
- naive bayes models
- high dimensional
- machine learning