On the Expressiveness of Approximate Inference in Bayesian Neural Networks.
Andrew Y. K. FoongDavid R. BurtYingzhen LiRichard E. TurnerPublished in: NeurIPS (2020)
Keyphrases
- approximate inference
- neural network
- bayesian networks
- graphical models
- probabilistic inference
- expectation propagation
- variational approximation
- belief propagation
- exact inference
- markov chain monte carlo
- parameter estimation
- message passing
- gaussian process
- latent variables
- loopy belief propagation
- variational methods
- variational inference
- posterior distribution
- dynamic bayesian networks
- factor graphs
- posterior probability
- artificial neural networks
- conditional random fields
- maximum likelihood
- bayesian inference
- probabilistic model
- structured prediction
- competitive learning
- multilayer perceptron
- model selection
- free energy
- graph cuts
- stereo matching
- bayesian methods
- neural network model