WAVELET ESTIMATION OF STATE SPACE MODELS

Airlane P. Alencar (Brazil), Pedro A. Morettin (Brazil), Clélia M. C. Toloi (Brazil) and Eliana Zandonade (Brazil)

Received January 21, 2009

Abstract

In this paper, we review recent work on state space switching models and add some new material on the estimation of parameters of a state space model through the use of the Kalman filter, the expectation-maximization (EM) algorithm and wavelets. We discuss specially switching models where the changes of regimes may be independent or following a Markov chain. We provide a simulation and an application to real data.

Keywords and phrases: EM algorithm, Kalman filter, Markov chain, state space models, switching models, wavelets.