Variational parameter learning in sequential state-space model via particle filtering
Accepted version
Peer-reviewed
Repository URI
Repository DOI
Change log
Authors
Li, Chenhao https://orcid.org/0000-0001-6731-3675
Godsill, SJ
Abstract
Parameter learning of the state-space model (SSM) plays a significant role in the modelling of time-series data and dynamical systems. However, the closed-form inference of the parameter posterior is often limited by sequential construction and non-linearity of the SSMs, which has led to the development of sampling-based algorithms such as particle Markov chain Monte Carlo (PMCMC). We present a novel algorithm, the particle filter variational inference (PF-VI) algorithm, which achieves closed-form learning of SSM parameters while tractably inferring the non-linear sequential states. We apply the algorithm to a popular non-linear SSM example and compare its performance against two competing PMCMC algorithms.
Description
Keywords
Bayesian inference, state-space model, variational Bayes, particle filtering
Journal Title
ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Conference Name
ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Journal ISSN
1520-6149
Volume Title
2021-June
Publisher
IEEE
Publisher DOI
Rights
All rights reserved