Informative Gaussian Scale Mixture Priors for Bayesian Neural Networks.
Tianyu CuiAki S. HavulinnaPekka MarttinenSamuel KaskiPublished in: CoRR (2020)
Keyphrases
- neural network
- heavy tailed
- generalized em algorithm
- image prior
- maximum likelihood
- prior distribution
- gaussian distribution
- mixture model
- mixture distribution
- gaussian mixture
- gaussian mixture model
- gaussian model
- generalized gaussian
- mixture distributions
- prior probabilities
- dirichlet prior
- bayesian approaches
- gaussian density
- image restoration
- maximum a posteriori
- bayesian models
- fuzzy logic
- posterior distribution
- normal distribution
- pattern recognition
- scale parameter
- mixture of gaussians
- prior model
- scale space
- neural network model
- gaussian densities
- bayesian networks
- neural nets
- akaike information criterion
- em algorithm
- genetic algorithm
- markov fields
- probabilistic model
- dirichlet process
- gaussian derivatives
- prior knowledge
- exponential family
- scale invariant
- probability distribution
- multilayer perceptron
- prior information
- back propagation
- parameter estimation
- markov random field
- expectation maximization