Mitigating the Effects of Non-Identifiability on Inference for Bayesian Neural Networks with Latent Variables.
Yaniv YacobyWeiwei PanFinale Doshi-VelezPublished in: J. Mach. Learn. Res. (2022)
Keyphrases
- latent variables
- posterior distribution
- neural network
- variational bayes
- variational approximation
- probabilistic model
- exact inference
- variational inference
- structured prediction
- bayesian inference
- factor graphs
- markov chain monte carlo
- random variables
- bayesian networks
- bayesian learning
- probabilistic graphical models
- topic models
- latent variable models
- prior knowledge
- approximate inference
- structural svms
- hidden variables
- graphical models
- probabilistic inference
- real valued
- observed variables
- max margin
- causal relationships
- latent space
- gaussian process
- pairwise
- generative model
- causal models
- causal discovery
- conditionally independent
- causal effects
- gaussian processes
- free energy
- gibbs sampling
- latent dirichlet allocation
- posterior probability
- knowledge representation
- image segmentation