Zero-variance control variates (ZV-CV) are a post-processing method to reduce the variance of Monte Carlo estimators of expectations using the deriva-tives of the log target. Once the derivatives are available, the only additional com-putational effort lies in solving a linear regression problem. Significant variance reductions have been achieved with this method in low dimensional examples, but the number of covariates in the regression rapidly increases with the dimension of the target. In this paper, we present compelling empirical evidence that the use of penalized regression techniques in the selection of high-dimensional control vari-ates provides performance gains over the classical least squares method. Another type of regularization based on using subsets of derivatives, or a priori regular-ization as we refer to it in this paper, is also proposed to reduce computational and storage requirements. Several examples showing the utility and limitations of regularized ZV-CV for Bayesian inference are given. The methods proposed in this paper are accessible through the R package ZVCV.

Regularized Zero-Variance Control Variates

Mira, A;
2023-01-01

Abstract

Zero-variance control variates (ZV-CV) are a post-processing method to reduce the variance of Monte Carlo estimators of expectations using the deriva-tives of the log target. Once the derivatives are available, the only additional com-putational effort lies in solving a linear regression problem. Significant variance reductions have been achieved with this method in low dimensional examples, but the number of covariates in the regression rapidly increases with the dimension of the target. In this paper, we present compelling empirical evidence that the use of penalized regression techniques in the selection of high-dimensional control vari-ates provides performance gains over the classical least squares method. Another type of regularization based on using subsets of derivatives, or a priori regular-ization as we refer to it in this paper, is also proposed to reduce computational and storage requirements. Several examples showing the utility and limitations of regularized ZV-CV for Bayesian inference are given. The methods proposed in this paper are accessible through the R package ZVCV.
2023
Bayesian inference; controlled thermodynamic integration; CTI; curse of dimensionality; Markov Chain Monte Carlo simulation; MCMC; Monte Carlo simulations; penalized regression; sequential Monte Carlo; SMC; Stein operator; variance reduction
South, Lf; Oates, Cj; Mira, A; Drovandi, C
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11383/2162398
 Attenzione

L'Ateneo sottopone a validazione solo i file PDF allegati

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 4
  • ???jsp.display-item.citation.isi??? 1
social impact