Global inducing point variational posteriors for Bayesian neural networks and deep Gaussian processes.
Sebastian W. OberLaurence AitchisonPublished in: ICML (2021)
Keyphrases
- gaussian processes
- gaussian process
- hyperparameters
- bayesian approaches
- bayesian inference
- bayesian methods
- gaussian process regression
- covariance function
- multi task learning
- posterior distribution
- relevance vector machine
- multi task
- preference learning
- gaussian process models
- posterior probability
- human pose estimation
- bayesian networks
- dirichlet process
- closed form
- expectation propagation
- text classification