Fast Conditional Mixing of MCMC Algorithms for Non-log-concave Distributions

MaRDI QIDQ6510627arXivFDO

Authors Xiang Cheng, Bo-Han Wang, Yusong Zhu

Published in arXiv

Abstract: MCMC algorithms offer empirically efficient tools for sampling from a target distribution $pi(x)proptoexp(-V(x))$. However, on the theory side, MCMC algorithms suffer from slow mixing rate when $pi(x)$ is non-log-concave. Our work examines this gap and shows that when Poincar'e-style inequality holds on a subset $mathcalX$ of the state space, the conditional distribution of MCMC iterates over $mathcalX$ mixes fast to the true conditional distribution. This fast mixing guarantee can hold in cases when global mixing is provably slow. We formalize the statement and quantify the conditional mixing rate. We further show that conditional mixing can have interesting implications for sampling from mixtures of Gaussians, parameter estimation for Gaussian mixture models and Gibbs-sampling with well-connected local minima.

This page was built for publication: Fast Conditional Mixing of MCMC Algorithms for Non-log-concave Distributions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6510627)