Order estimation for a special class of hidden Markov sources and binary renewal processes

作者: S. Khudanpur , P. Narayan

DOI: 10.1109/TIT.2002.1003850

关键词:

摘要: We consider the estimation of order, i.e., number hidden states, a special class discrete-time finite-alphabet Markov sources. This can be characterized in terms equivalent renewal processes. No priori bound is assumed on maximum. permissible order. An order estimator based types constructed, and shown to strongly consistent by computing precise asymptotics probability error. The underestimation true decays exponentially observations while overestimation goes zero sufficiently fast. It further that this has best possible error exponent large estimators. Our results are also valid for general binary independent-renewal processes with finite mean times.

参考文章(20)
Brian G. Leroux, Maximum-likelihood estimation for hidden Markov models Stochastic Processes and their Applications. ,vol. 40, pp. 127- 143 ,(1992) , 10.1016/0304-4149(92)90141-C
Abraham J. Wyner, Peter B�hlmann, Variable length Markov chains Annals of Statistics. ,vol. 27, pp. 480- 513 ,(1999) , 10.1214/AOS/1018031204
N. Merhav, The estimation of the model order in exponential families IEEE Transactions on Information Theory. ,vol. 35, pp. 1109- 1114 ,(1989) , 10.1109/18.42231
G. H. Hardy, S. Ramanujan, Asymptotic Formulaae in Combinatory Analysis Proceedings of the London Mathematical Society. ,vol. s2-17, pp. 75- 115 ,(1918) , 10.1112/PLMS/S2-17.1.75
Erwin Bolthauser, Amir Dembo, Ofer Zeitouni, Large Deviation Techniques and Applications. Journal of the American Statistical Association. ,vol. 89, pp. 1145- ,(1994) , 10.2307/2290955
A. Barron, J. Rissanen, Bin Yu, The minimum description length principle in coding and modeling IEEE Transactions on Information Theory. ,vol. 44, pp. 2743- 2760 ,(1998) , 10.1109/18.720554
I. Csiszar, P.C. Shields, The consistency of the BIC Markov order estimator international symposium on information theory. pp. 26- ,(2000) , 10.1109/ISIT.2000.866316
J. Ziv, N. Merhav, Estimating the number of states of a finite-state source IEEE Transactions on Information Theory. ,vol. 38, pp. 61- 65 ,(1992) , 10.1109/18.108249