Title :
Adaptive estimation of hidden nearly completely decomposable Markov chains
Author :
Krishnamurthy, Vikram
Author_Institution :
Dept. of Syst. Eng., Australian Nat. Univ., Canberra, ACT, Australia
Abstract :
We propose maximum-likelihood (ML) estimation schemes for nearly completely decomposable Markov chains (NCDMC) in white Gaussian noise. Aggregation techniques based on stochastic complementation are applied to reduce the dimension of the resulting hidden Markov model (HMM) and hence substantially reduce the computational costs of the estimation algorithms. We then present an aggregation based expectation maximization (EM) algorithm for estimating the parameters and states of the HMM
Keywords :
FIR filters; Gaussian noise; adaptive equalisers; adaptive estimation; adaptive signal processing; hidden Markov models; maximum likelihood estimation; optimisation; state estimation; white noise; adaptive estimation; aggregation techniques; computational costs; estimation algorithms; expectation maximization algorithm; hidden Markov model; hidden nearly completely decomposable Markov chains; maximum-likelihood estimation schemes; stochastic complementation; white Gaussian noise; Adaptive estimation; Australia; Computational efficiency; Gaussian noise; Hidden Markov models; Maximum likelihood estimation; Parameter estimation; State estimation; Stochastic processes; Systems engineering and theory;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1994. ICASSP-94., 1994 IEEE International Conference on
Conference_Location :
Adelaide, SA
Print_ISBN :
0-7803-1775-0
DOI :
10.1109/ICASSP.1994.389811
