Fast compression of MCMC output

07/09/2021
by   Nicolas Chopin, et al.
0

We propose cube thinning, a novel method for compressing the output of a MCMC (Markov chain Monte Carlo) algorithm when control variates are available. It amounts to resampling the initial MCMC sample (according to weights derived from control variates), while imposing equality constraints on averages of these control variates, using the cube method of <cit.>. Its main advantage is that its CPU cost is linear in N, the original sample size, and is constant in M, the required size for the compressed sample. This compares favourably to Stein thinning <cit.>, which has complexity 𝒪(NM^2), and which requires the availability of the gradient of the target log-density (which automatically implies the availability of control variates). Our numerical experiments suggest that cube thinning is also competitive in terms of statistical error.

READ FULL TEXT

page 15

page 16

page 17

page 18

research
05/23/2020

Most Likely Optimal Subsampled Markov Chain Monte Carlo

Markov Chain Monte Carlo (MCMC) requires to evaluate the full data likel...
research
05/08/2020

Optimal Thinning of MCMC Output

The use of heuristics to assess the convergence and compress the output ...
research
10/23/2020

No Free Lunch for Approximate MCMC

It is widely known that the performance of Markov chain Monte Carlo (MCM...
research
04/16/2014

Speeding Up MCMC by Efficient Data Subsampling

We propose Subsampling MCMC, a Markov Chain Monte Carlo (MCMC) framework...
research
01/10/2021

The shifted ODE method for underdamped Langevin MCMC

In this paper, we consider the underdamped Langevin diffusion (ULD) and ...
research
03/28/2014

Accelerating MCMC via Parallel Predictive Prefetching

We present a general framework for accelerating a large class of widely ...
research
10/30/2016

Auxiliary gradient-based sampling algorithms

We introduce a new family of MCMC samplers that combine auxiliary variab...

Please sign up or login with your details

Forgot password? Click here to reset