Identifiability of deep generative models under mixture priors without auxiliary information.
Bohdan KivvaGoutham RajendranPradeep RavikumarBryon AragamPublished in: CoRR (2022)
Keyphrases
- generative model
- auxiliary information
- mixture model
- prior knowledge
- bayesian framework
- expectation maximization
- probabilistic model
- external knowledge
- deep belief networks
- em algorithm
- discriminative learning
- discriminative models
- latent dirichlet allocation
- generative and discriminative models
- deep learning
- domain knowledge
- topic models
- semi supervised
- image segmentation
- maximum likelihood
- object recognition
- training data