Deep neural networks with dependent weights: Gaussian Process mixture limit, heavy tails, sparsity and compressibility.
Hoil LeeFadhel AyedPaul JungJuho LeeHongseok YangFrançois CaronPublished in: CoRR (2022)
Keyphrases
- gaussian process
- heavy tails
- neural network
- heavy tailed
- sparse approximations
- gaussian processes
- regression model
- bayesian framework
- model selection
- semi supervised
- hyperparameters
- mixture model
- latent variables
- linear combination
- sparse representation
- generalized gaussian
- probability density function
- high dimensional
- expectation maximization
- feature space
- machine learning
- unsupervised learning
- density estimation
- maximum likelihood
- active learning
- gaussian model
- pairwise