Kernel Stein Discrepancy thinning: a theoretical perspective of pathologies and a practical fix with regularization - ENSAE Paris Access content directly
Preprints, Working Papers, ... Year : 2023

Kernel Stein Discrepancy thinning: a theoretical perspective of pathologies and a practical fix with regularization

Abstract

Stein thinning is a promising algorithm proposed by (Riabiz et al., 2022) for post-processing outputs of Markov chain Monte Carlo (MCMC). The main principle is to greedily minimize the kernelized Stein discrepancy (KSD), which only requires the gradient of the log-target distribution, and is thus well-suited for Bayesian inference. The main advantages of Stein thinning are the automatic remove of the burn-in period, the correction of the bias introduced by recent MCMC algorithms, and the asymptotic properties of convergence towards the target distribution. Nevertheless, Stein thinning suffers from several empirical pathologies, which may result in poor approximations, as observed in the literature. In this article, we conduct a theoretical analysis of these pathologies, to clearly identify the mechanisms at stake, and suggest improved strategies. Then, we introduce the regularized Stein thinning algorithm to alleviate the identified pathologies. Finally, theoretical guarantees and extensive experiments show the high efficiency of the proposed algorithm.
Fichier principal
Vignette du fichier
main.pdf (1.83 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03962614 , version 1 (31-01-2023)
hal-03962614 , version 2 (17-05-2023)
hal-03962614 , version 3 (25-10-2023)

Identifiers

Cite

Clément Bénard, Brian Staber, Sébastien Da Veiga. Kernel Stein Discrepancy thinning: a theoretical perspective of pathologies and a practical fix with regularization. 2023. ⟨hal-03962614v1⟩
114 View
79 Download

Altmetric

Share

Gmail Facebook X LinkedIn More