AdamB: Decoupled Bayes by Backprop With Gaussian Scale Mixture Prior.
Keigo NishidaMakoto TaijiPublished in: IEEE Access (2022)
Keyphrases
- gaussian distribution
- gaussian mixture model
- gaussian mixture
- mixture distribution
- gaussian model
- gaussian densities
- mixture model
- scale space
- maximum likelihood
- covariance matrices
- gaussian derivatives
- mixture of gaussians
- decision trees
- gaussian density
- small scale
- support vector
- maximum a posteriori
- zero crossing
- multivariate gaussian
- prior knowledge
- expectation maximization
- scale parameter
- heavy tailed
- mixture distributions
- image processing
- generalized gaussian
- bayes classifier
- multiscale
- bayesian model
- feature space
- closed form