Differentiable Bootstrap Particle Filters for Regime-Switching Models

TL;DR

This paper introduces a differentiable particle filter capable of learning and tracking regime-switching models, effectively handling dynamic changes in state and measurement models in complex real-world scenarios.

Contribution

It presents a novel differentiable particle filter that learns multiple candidate models and tracks states in regime-switching environments, advancing model adaptability.

Findings

Outperforms existing algorithms in regime-switching scenarios

Successfully learns unknown dynamic and measurement models

Demonstrates robustness in complex real-world applications

Abstract

Differentiable particle filters are an emerging class of particle filtering methods that use neural networks to construct and learn parametric state-space models. In real-world applications, both the state dynamics and measurements can switch between a set of candidate models. For instance, in target tracking, vehicles can idle, move through traffic, or cruise on motorways, and measurements are collected in different geographical or weather conditions. This paper proposes a new differentiable particle filter for regime-switching state-space models. The method can learn a set of unknown candidate dynamic and measurement models and track the state posteriors. We evaluate the performance of the novel algorithm in relevant models, showing its great performance compared to other competitive algorithms.