Abstract
It is shown that there does not exist a complete sufficient statistic for the class of allk-state Markov chains,k≧3. Whenk=2 there is a complete sufficient statistic.
Similar content being viewed by others
Liferences
Billingsley, P. (1961). Statistical methods in Markov chains,Ann. Math. Statist.,32, 12–40.
Denny, J. L. and Wright, A. L. (1978). On tests for Markov dependence.Zeit. Wahr.-scheinlichkeitsh.,43, 331–338.
Denny, J. L. and Yakowitz, S. J. (1978). Admissible run-contingency type tests for independence and Markov dependence,J. Amer. Statist. Ass.,73, 177–181.
Lehmann, E. L. (1959).Testing Statistical Hypotheses, Wiley, New York.
Lehmann, E. L. and Scheffé, H. (1950, 1955). Completeness, similar regions, and unbiased estimation.Sankhyà,10, 305–340 andSankhyà,15, 219–236.
Zacks, S. (1971).The Theory of Statistical Inference, Wiley, New York.
Author information
Authors and Affiliations
About this article
Cite this article
Wright, A.L. Nonexistence of complete sufficient statistics for stationaryk-state Markov chains,k≧3. Ann Inst Stat Math 32, 95–97 (1980). https://doi.org/10.1007/BF02480314
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1007/BF02480314