Fiducial Matching for the Approximate Posterior: F-ABC

TL;DR

F-ABC introduces a novel approach using universal sufficient statistics for approximate Bayesian computation, avoiding kernel artifacts and providing better asymptotic properties, with practical guidelines for tolerance selection.

Contribution

It presents F-ABC, a new ABC method leveraging universal sufficient statistics, improving accuracy and robustness over previous kernel-based approaches.

Findings

F-ABC avoids artifacts caused by kernels in ABC.

Asymptotic analysis supports the method's validity.

Simulation results demonstrate improved performance.

Abstract

F-ABC is introduced, using universal sufficient statistics, unlike previous ABC papers, e.g. Bernton et al. (2019), and avoiding in the approximate posterior artifacts due to a Kernel. The nature of matching tolerance is examined and indications for determining its values are presented. F-ABC does not face concerns associated with ABC. Asymptotics and simulation results are also presented.