Research Repository

Bayesian Fusion

Dai, Hongsheng and Pollock, Murray and Roberts, Gareth (2019) Bayesian Fusion. Working Paper. under review. (Submitted)

BayesianFusion.pdf - Submitted Version

Download (492kB) | Preview


The paper presents an exact Bayesian Fusion algorithm, which can carry out perfect inferences for the unification of distributed data analysis. The new method uses parallel but coalesced Markov processes to drive distributed Monte Carlo draws to a Monte Carlo sample from the posterior of the full data. The Markov processes are simulated via path-space rejection sampling for diffusion processes. The methodology of this exact Bayesian Fusion algorithm explains why existing methods do not provide good results and how to correct approximated draws of existing methods in order to obtain exact samples. Its approximate version, the sequential Bayesian Fusion algorithm, can be implemented in parallel for big data analysis. The sequential Bayesian Fusion outperforms all existing methods, which is justified theoretically and via numerical studies.

Item Type: Monograph (Working Paper)
Uncontrolled Keywords: Big Data, Distributed Analysis, Parallel Computation, Path-space Rejection Sampling, Sequential Monte Carlo, Unification Monte Carlo
Divisions: Faculty of Science and Health > Mathematical Sciences, Department of
Depositing User: Elements
Date Deposited: 20 Nov 2019 11:56
Last Modified: 20 Nov 2019 12:15

Actions (login required)

View Item View Item