A convergent scheme for the Bayesian filtering problem based on the Fokker--Planck equation and deep splitting

TL;DR

This paper introduces a deep splitting-based numerical scheme for nonlinear Bayesian filtering that approximates the Fokker--Planck equation, with proven convergence and demonstrated robustness in high-dimensional examples.

Contribution

It presents a novel deep splitting algorithm for Bayesian filtering with theoretical convergence guarantees and practical effectiveness in high-dimensional settings.

Findings

Convergence rate established under Hörmander condition.

Algorithm performs robustly in a 10-dimensional numerical example.

Mitigates curse of dimensionality in filtering problems.

Abstract

A numerical scheme for approximating the nonlinear filtering density is introduced and its convergence rate is established, theoretically under a parabolic H\"{o}rmander condition, and empirically in numerical examples. In a prediction step, between the noisy and partial measurements at discrete times, the scheme approximates the Fokker--Planck equation with a deep splitting scheme, followed by an exact update through Bayes' formula. This results in a classical prediction-update filtering algorithm that operates online for new observation sequences post-training. The algorithm employs a sampling-based Feynman--Kac approach, designed to mitigate the curse of dimensionality. As a corollary we obtain the convergence rate for the approximation of the Fokker--Planck equation alone, disconnected from the filtering problem. The convergence analysis is complemented by a nonlinear $10$-dimensional numerical example demonstrating the robustness of the method.