Bayesian Inference in Recurrent Explicit Duration Switching Linear Dynamical Systems
This work addresses segmentation challenges in dynamical systems, likely for applications in time-series analysis, but appears incremental as it builds upon existing rSLDS models.
The authors tackled the problem of improving segmentation in switching linear dynamical systems by proposing a new model called REDSLDS that incorporates recurrent explicit duration variables, and they demonstrated enhanced segmentation capabilities on three benchmark datasets.
In this paper, we propose a novel model called Recurrent Explicit Duration Switching Linear Dynamical Systems (REDSLDS) that incorporates recurrent explicit duration variables into the rSLDS model. We also propose an inference and learning scheme that involves the use of Pólya-gamma augmentation. We demonstrate the improved segmentation capabilities of our model on three benchmark datasets, including two quantitative datasets and one qualitative dataset.