End-to-End Identifiable and Consistent Recurrent Switching Dynamical Systems

TL;DR

This paper introduces a new flow-based estimator for recurrent switching dynamical systems that improves identifiability and disentanglement over traditional VAE methods, with strong theoretical and empirical results.

Contribution

It establishes broad identifiability results for nonlinear switching systems and proposes ΩSDS, a flow-based estimator enabling exact likelihood optimization.

Findings

ΩSDS achieves better disentanglement than VAE-based estimators.

Theoretical extension of identifiability to more general nonlinear systems.

Empirical results show improved forecasting accuracy.

Abstract

Learning identifiable representations in deep generative models remains a fundamental challenge, particularly for sequential data with regime-switching dynamics. Existing approaches establish identifiability under restrictive assumptions, such as stationarity or limited emission models, and typically rely on variational autoencoder (VAE) estimators, which introduce approximation gaps that limit the recovery of the latent structure. In this work, we address both the theoretical and practical limitations of this setting. First, we establish identifiability of a broad class of recurrent nonlinear switching dynamical systems under flexible assumptions, significantly extending prior results. Second, we introduce $\Omega$SDS, a flow-based estimator that enables exact likelihood optimization using expectation-maximisation. Through empirical validation on both synthetic and real-world data, our results demonstrate that $\Omega$SDS achieves improved disentanglement compared to VAE-based estimators and more accurate forecasting of underlying dynamics.