Sequential Monte Carlo with Adaptive Weights for Approximate Bayesian Computation

TL;DR

This paper introduces an adaptive weights extension to sequential Monte Carlo methods for approximate Bayesian computation, significantly improving acceptance rates in complex model analysis.

Contribution

The paper presents a simple, computationally trivial adaptive weights approach for ABC SMC, enhancing efficiency over existing methods.

Findings

Substantially improved acceptance rates demonstrated in simulated data

Effective application to real data in systems biology models

Easy implementation with minimal computational overhead

Abstract

Methods of approximate Bayesian computation (ABC) are increasingly used for analysis of complex models. A major challenge for ABC is over-coming the often inherent problem of high rejection rates in the accept/reject methods based on prior:predictive sampling. A number of recent developments aim to address this with extensions based on sequential Monte Carlo (SMC) strategies. We build on this here, introducing an ABC SMC method that uses data-based adaptive weights. This easily implemented and computationally trivial extension of ABC SMC can very substantially improve acceptance rates, as is demonstrated in a series of examples with simulated and real data sets, including a currently topical example from dynamic modelling in systems biology applications.