Temporal Parallelization of Bayesian Smoothers
Computation
2020-02-21 v2 Distributed, Parallel, and Cluster Computing
Dynamical Systems
Abstract
This paper presents algorithms for temporal parallelization of Bayesian smoothers. We define the elements and the operators to pose these problems as the solutions to all-prefix-sums operations for which efficient parallel scan-algorithms are available. We present the temporal parallelization of the general Bayesian filtering and smoothing equations and specialize them to linear/Gaussian models. The advantage of the proposed algorithms is that they reduce the linear complexity of standard smoothing algorithms with respect to time to logarithmic.
Cite
@article{arxiv.1905.13002,
title = {Temporal Parallelization of Bayesian Smoothers},
author = {Simo Särkkä and Ángel F. García-Fernández},
journal= {arXiv preprint arXiv:1905.13002},
year = {2020}
}