Constraining Variational Inference with Geometric Jensen-Shannon Divergence.
Jacob DeasyNikola SimidjievskiPietro LióPublished in: NeurIPS (2020)
Keyphrases
- jensen shannon divergence
- variational inference
- bayesian inference
- information theory
- information theoretic
- posterior distribution
- probabilistic model
- probabilistic graphical models
- mixture model
- topic models
- gaussian process
- feature selection
- variational methods
- latent dirichlet allocation
- closed form
- mutual information
- approximate inference
- exact inference
- selection criterion
- exponential family
- computer vision
- hyperparameters
- probabilistic inference
- bayesian framework
- prior information
- generative model
- graphical models
- data sets
- latent variables
- factor graphs
- text mining