AUTOREGRESSIVE MODEL WITH PARTIAL FORGETTING WITHIN RAO-BLACKWELLIZED PARTICLE FILTER

Kamil Dedecius (1) , Radek Hofman (1)

1 Department of Adaptive Systems

We are concerned with Bayesian identification and prediction of a nonlinear discrete stochastic process. The fact, that a nonlinear process can be approximated by a piecewise linear function advocates the use of adaptive linear models. We propose a linear regression model within Rao-Blackwellized particle filter. The parameters of the linear model are adaptively estimated using a finite mixture, where the weights of components are tuned with a particle filter. The mixture reflects a priori given hypotheses on different scenarios of (expected) parameters' evolution. The resulting hybrid filter locally optimizes the weights to achieve the best fit of signal.

Mots clés: Physical Sciences

Domaines: Calcul [stat.CO]

Kamil Dedecius, Radek Hofman. AUTOREGRESSIVE MODEL WITH PARTIAL FORGETTING WITHIN RAO-BLACKWELLIZED PARTICLE FILTER. Communications in Statistics - Simulation and Computation, 2011, 41 (05), pp.582-589. ⟨10.1080/03610918.2011.598992⟩. ⟨hal-00768970⟩