Metropolis-Hastings view on variational inference and adversarial training.
Kirill NeklyudovPavel ShvechikovDmitry P. VetrovPublished in: CoRR (2018)
Keyphrases
- variational inference
- posterior distribution
- variational approximation
- bayesian inference
- topic models
- mixture model
- probabilistic graphical models
- gaussian process
- latent dirichlet allocation
- probability distribution
- probabilistic model
- parameter estimation
- graphical models
- closed form
- maximum a posteriori
- latent variables
- exponential family
- markov chain monte carlo
- variational methods
- variational bayes
- incremental learning
- exact inference
- machine learning
- bayesian framework
- higher order
- training set