Scalable Bayesian Learning for State Space Models using Variational Inference with SMC Samplers

Abstract

We present a scalable approach to performing approximate fully Bayesian inference in generic state space models. The proposed method is an alternative to particle MCMC that provides fully Bayesian inference of both the dynamic latent states and the static parameters of the model. We build up on recent advances in computational statistics that combine variational methods with sequential Monte Carlo sampling and we demonstrate the advantages of performing full Bayesian inference over the static parameters rather than just performing variational EM approximations. We illustrate how our approach enables scalable inference in multivariate stochastic volatility models and self-exciting point process models that allow for flexible dynamics in the latent intensity function.

arXiv commentTo appear in AISTATS 2019

Cited by in corpus (5)

• Differentiable Particle Filtering via Entropy-Regularized Optimal Transport
• The Neural Moving Average Model for Scalable Variational Inference of State Space Models
• Fast and Flexible Temporal Point Processes with Triangular Maps
• Relational State-Space Model for Stochastic Multi-Object Systems
• Variational Dynamic Mixtures