A Bayesian Approach to Sparse plus Low rank Network Identification

TL;DR

This paper introduces a Bayesian method for identifying sparse plus low rank models in multivariate time series, enabling efficient modeling of complex dynamical systems with latent variables.

Contribution

It proposes a Gaussian regression approach to accurately identify sparse plus low rank structures in multivariate time series models.

Findings

Effective identification of sparse plus low rank models

Improved modeling of latent variables in time series

Enhanced interpretability of dynamical networks

Abstract

We consider the problem of modeling multivariate time series with parsimonious dynamical models which can be represented as sparse dynamic Bayesian networks with few latent nodes. This structure translates into a sparse plus low rank model. In this paper, we propose a Gaussian regression approach to identify such a model.