A Filtered EM Algorithm for Parameter Estimation in Linear Filtering
Patrick Florchinger · Birkhäuser Basel eBooks · 2002
The aim of this paper is to compute finite-dimensional filters for integrals and stochastic integrals of moments of the state of partially observed linear Gaussian systems when the signal process and the observation noise are possibly correlated. These filters are used with the expectation maximization (EM) algorithm in order to obtain maximum likelihood estimation of the system parameters.