Filtering Jump Markov Systems with Partially Known Dynamics: A Model-Based Deep Learning Approach

TL;DR

This paper introduces JMFNet, a deep learning framework combining RNNs for mode prediction and filtering, enabling real-time state estimation in jump Markov systems with unknown dynamics, outperforming classical and model-free methods.

Contribution

It proposes a hybrid RNN-based architecture trained jointly for mode prediction and filtering, handling unknown noise and transition dynamics without supervision.

Findings

Outperforms classical filters like IMM and particle filters.

Achieves small but meaningful improvements over KalmanNet.

Demonstrates robustness in non-stationary, high-noise, and complex systems.

Abstract

This paper presents the Jump Markov Filtering Network (JMFNet), a novel model-based deep learning framework for real-time state-state estimation in jump Markov systems with unknown noise statistics and mode transition dynamics. A hybrid architecture comprising two Recurrent Neural Networks (RNNs) is proposed: one for mode prediction and another for filtering that is based on a mode-augmented version of the recently presented KalmanNet architecture. The proposed RNNs are trained jointly using an alternating least squares strategy that enables mutual adaptation without supervision of the latent modes. Extensive numerical experiments on linear and nonlinear systems, including target tracking, pendulum angle tracking, Lorenz attractor dynamics, and a real-life dataset demonstrate that the proposed JMFNet framework outperforms classical model-based filters (e.g., interacting multiple models and particle filters) as well as model-free deep learning baselines, particularly in non-stationary and high-noise regimes. It is also showcased that JMFNet achieves a small yet meaningful improvement over the KalmanNet framework, which becomes much more pronounced in complicated systems or long trajectories. Finally, the method's performance is empirically validated to be consistent and reliable, exhibiting low sensitivity to initial conditions, hyperparameter selection, as well as to incorrect model knowledge