Loss-Calibrated Approximate Inference in Bayesian Neural Networks.
Adam D. CobbStephen J. RobertsYarin GalPublished in: CoRR (2018)
Keyphrases
- belief networks
- approximate inference
- graphical models
- probabilistic inference
- neural network
- exact inference
- belief propagation
- bayesian networks
- variational approximation
- multilayer perceptron
- markov chain monte carlo
- conditional probabilities
- expectation propagation
- probabilistic model
- variational methods
- variational inference
- loopy belief propagation
- factor graphs
- probability distribution
- message passing
- random variables
- conditional random fields
- bayesian inference
- artificial neural networks
- dynamic bayesian networks
- posterior distribution
- structured prediction
- parameter estimation
- multi view
- free energy
- posterior probability
- latent variables
- neural network model
- maximum likelihood
- least squares
- competitive learning
- three dimensional
- image processing