Adaptive sequential Monte Carlo for multiple changepoint analysis

TL;DR

This paper presents an efficient, adaptive sequential Monte Carlo algorithm for real-time detection of multiple changepoints in continuous data streams, improving accuracy and computational efficiency over existing methods.

Contribution

It introduces a simple, adaptive SMC method that updates changepoint estimates based on recent data, with dynamic particle allocation for better performance and efficiency.

Findings

Outperforms current state-of-the-art changepoint detection methods.

Adaptive algorithm reduces computation when changes are negligible.

Effective in inferring Poisson process intensities with Bayesian models.

Abstract

Process monitoring and control requires detection of structural changes in a data stream in real time. This article introduces an efficient sequential Monte Carlo algorithm designed for learning unknown changepoints in continuous time. The method is intuitively simple: new changepoints for the latest window of data are proposed by conditioning only on data observed since the most recent estimated changepoint, as these carry most of the information about the state of the process prior to the update. The proposed method shows improved performance over the current state of the art. Another advantage of the proposed algorithm is that it can be made adaptive, varying the number of particles according to the apparent local complexity of the target changepoint probability distribution. This saves valuable computing time when changes in the change- point distribution are negligible, and enables re-balancing of the importance weights of ex- isting particles when a significant change in the target distribution is encountered. The plain and adaptive versions of the method are illustrated using the canonical con- tinuous time changepoint problem of inferring the intensity of an inhomogeneous Poisson process. Performance is demonstrated using both conjugate and non-conjugate Bayesian models for the intensity.