Current browse context:
stat
Change to browse by:
References & Citations
Statistics > Computation
Title: Merging MCMC Subposteriors through Gaussian-Process Approximations
(Submitted on 27 May 2016 (v1), last revised 17 Jul 2017 (this version, v2))
Abstract: Markov chain Monte Carlo (MCMC) algorithms have become powerful tools for Bayesian inference. However, they do not scale well to large-data problems. Divide-and-conquer strategies, which split the data into batches and, for each batch, run independent MCMC algorithms targeting the corresponding subposterior, can spread the computational burden across a number of separate workers. The challenge with such strategies is in recombining the subposteriors to approximate the full posterior. By creating a Gaussian-process approximation for each log-subposterior density we create a tractable approximation for the full posterior. This approximation is exploited through three methodologies: firstly a Hamiltonian Monte Carlo algorithm targeting the expectation of the posterior density provides a sample from an approximation to the posterior; secondly, evaluating the true posterior at the sampled points leads to an importance sampler that, asymptotically, targets the true posterior expectations; finally, an alternative importance sampler uses the full Gaussian-process distribution of the approximation to the log-posterior density to re-weight any initial sample and provide both an estimate of the posterior expectation and a measure of the uncertainty in it.
Submission history
From: Christopher Nemeth [view email][v1] Fri, 27 May 2016 10:51:48 GMT (202kb,D)
[v2] Mon, 17 Jul 2017 14:35:44 GMT (412kb,D)
Link back to: arXiv, form interface, contact.