conftrace_

2016 ICML ICML 2016

Mixing Rates for the Alternating Gibbs Sampler over Restricted Boltzmann Machines and Friends

Abstract

Alternating Gibbs sampling is a modification of classical Gibbs sampling where several variables are simultaneously sampled from their joint conditional distribution. In this work, we investigate the mixing rate of alternating Gibbs sampling with a particular emphasis on Restricted Boltzmann Machines (RBMs) and variants.

Authors

Christopher Tosh

Topics

Machine Learning > Optimization & Theory > Bayesian Inference Machine Learning > Optimization & Theory > Stochastic Processes Deep Learning > Architectures > Neural Networks

Keywords

markov chain monte carlo gibbs sampling restricted boltzmann machine mixing rate alternating gibbs sampler

More from ICML 2016

All ICML 2016 papers →

Online Learning with Feedback Graphs Without the Graphs

PAC learning of Probabilistic Automaton based on the Method of Moments

Estimating Structured Vector Autoregressive Models

Polynomial Networks and Factorization Machines: New Insights and Efficient Training Algorithms

A New PAC-Bayesian Perspective on Domain Adaptation

Correlation Clustering and Biclustering with Locally Bounded Errors