Automatic Rao-Blackwellization for Sequential Monte Carlo with Belief Propagation

Waïss Azizian, Guillaume Baudart, Marc Lelarge

Keywords: Kalman filter, probabilistic inference, reactive programming, graphical models

TL;DR: A new Sequential Monte Carlo algorithm mixing exact computation and sampling automatically.

Abstract:

Exact Bayesian inference on state-space models (SSM) is in general untractable and, unfortunately, basic Sequential Monte Carlo (SMC) methods do not yield correct approximations for complex models. In this paper, we propose a mixed inference algorithm that computes closed-form solutions using Belief Propagation as much as possible, and falls back to sampling-based SMC methods when exact computations fail. This algorithm thus implements automatic Rao-Blackwellization and is even exact for Gaussian tree models.

Submission Number: 74