DocumentCode :
1506509
Title :
New finite-dimensional filters for parameter estimation of discrete-time linear Gaussian models
Author :
Elliott, Robert J. ; Krishnamurthy, Vikram
Author_Institution :
Dept. of Math. Sci., Alberta Univ., Edmonton, Alta., Canada
Volume :
44
Issue :
5
fYear :
1999
fDate :
5/1/1999 12:00:00 AM
Firstpage :
938
Lastpage :
951
Abstract :
The authors derive a new class of finite-dimensional recursive filters for linear dynamical systems. The Kalman filter is a special case of their general filter. Apart from being of mathematical interest, these new finite-dimensional filters can be used with the expectation maximization (EM) algorithm to yield maximum likelihood estimates of the parameters of a linear dynamical system. Important advantages of their filter-based EM algorithm compared with the standard smoother-based EM algorithm include: 1) substantially reduced memory requirements, and 2) ease of parallel implementation on a multiprocessor system. The algorithm has applications in multisensor signal enhancement of speech signals and also econometric modeling
Keywords :
Gaussian processes; Kalman filters; discrete time systems; filtering theory; linear systems; maximum likelihood estimation; recursive filters; state-space methods; Kalman filter; discrete-time systems; econometric modeling; expectation maximization algorithm; finite-dimensional filters; linear Gaussian models; linear dynamical systems; maximum likelihood estimation; multisensor signal enhancement; parameter estimation; recursive filters; speech processing; state space model; Hidden Markov models; Information processing; Iterative algorithms; Maximum likelihood estimation; Multiprocessing systems; Nonlinear filters; Parameter estimation; Recursive estimation; Signal processing; Statistics;
fLanguage :
English
Journal_Title :
Automatic Control, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9286
Type :
jour
DOI :
10.1109/9.763210
Filename :
763210
Link To Document :
بازگشت