Name
Title | ||
|---|---|---|
On-line estimation of hidden Markov model parameters based on the Kullback-Leibler information measure |
Abstract | ||
|---|---|---|
Inthis paper, sequential or "on-line" hidden Mar- kov model (HMM) signal processing schemes are derived and their performance illustrated in simulation studies. The on-line algorithms are sequential expectation maximization (EM) schemes and are derived by using stochastic approximations to maximize the Kullback-Leibler information measure. The whemes can be implemented either as filters or fixed-lag or .wtooth-lag smoothers. They yield estimates of the HMM pa- ;meters including transition probabilities, Markov state lev- cis, and noise variance. In contrast to the off-line EM algorithm (Baum Welch scheme) which uses the fixed-interval "forward- backward" scheme, the on-line schemes have significantly re- duced memory requirements, improved convergence (as shown in simulations), and can estimate HMM parameters that vary slowly with time or undergo infrequent jump changes. Using similar techniques we also derive on-line schemes to extract finite-state Markov chains imbedded in a mixture of white Gaussian noise (WGN) and deterministic signals of known ;',lnctional form with unknown parameters, In particular, de- :rministic periodic signals with unknown and time-varying amplitudes and phases are considered. Simulations presented show that these schemes satisfactorily estimate the HMM pa- rameters and also the time-varying amplitudes and phases. |
Year | DOI | Venue |
|---|---|---|
1993 | 10.1109/78.229888 | Signal Processing, IEEE Transactions |
Keywords | Field | DocType |
filtering and prediction theory,hidden Markov models,parameter estimation,signal processing,white noise,EM algorithm,HMM,Kullback-Leibler information measure,Markov state levels,convergence,deterministic signals,filters,finite-state Markov chains,fixed-lag smoothers,hidden Markov model,noise variance,online algorithms,parameter estimation,sawtooth-lag smoothers,sequential expectation maximization,signal processing,stochastic approximations,transition probabilities,white Gaussian noise | Markov process,Pattern recognition,Markov model,Expectation–maximization algorithm,Markov chain,White noise,Artificial intelligence,Estimation theory,Hidden Markov model,Gaussian noise,Mathematics | Journal |
Volume | Issue | ISSN |
41 | 8 | 1053-587X |
Citations | PageRank | References |
125 | 19.16 | 5 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Vikram Krishnamurthy | 1 | 925 | 162.74 |
| JOHN B. MOORE | 2 | 412 | 84.61 |