On Sequential Monte Carlo Sampling Methods for Bayesian Filtering

Abstract

Abstract Many problems in applied statistics, statistical signal processing, time series analysis and econometrics can be stated in a state space form as follows. A transition equation describes the prior distribution of a hidden Markov process xk  k N , the so-called hidden state process, and an observation equation describes the likelihood of the observations yk  k N ,k being a discrete time index.