Bayesian System Identification of Dynamical Systems using Reversible Jump Markov Chain Monte Carlo

摘要

The purpose of this contribution is to illustrate the potential of Reversible Jump Markov Chain Monte Carloud(RJMCMC) methods for nonlinear system identification. Markov Chain Monte Carlo (MCMC) samplingudmethods have come to be viewed as a standard tool for tackling the issue of parameter estimation usingudBayesian inference. A limitation of standard MCMC approaches is that they are not suited to tackling theudissue of model selection. RJMCMC offers a powerful extension to standard MCMC approaches in that itudallows parameter estimation and model selection to be addressed simultaneously. This is made possibleudby the fact that the RJMCMC algorithm is able to jump between parameter spaces of varying dimension.udIn this paper the background theory to the RJMCMC algorithm is introduced. Comparison is made to audstandard MCMC approach.