Efficient MCMC Sampling for Bayesian Matrix Factorization by Breaking Posterior Symmetries.
Saibal DeHadi SalehiAlex A. GorodetskyPublished in: CoRR (2020)
Keyphrases
- matrix factorization
- markov chain monte carlo
- posterior distribution
- posterior probability
- variational bayesian
- monte carlo
- collaborative filtering
- bayesian inference
- exponential family
- sampling algorithm
- parameter estimation
- metropolis hastings
- metropolis hastings algorithm
- markov chain
- particle filter
- probability distribution
- factorization methods
- low rank
- recommender systems
- proposal distribution
- importance sampling
- factor analysis
- negative matrix factorization
- approximate inference
- generative model
- dirichlet process
- particle filtering
- bayesian framework
- probabilistic matrix factorization
- latent variables
- stochastic gradient descent
- missing data
- maximum likelihood
- natural images
- hyperparameters