VECTOR MARKOV CHAIN WITH PARTIAL CONNECTIONS AND STATISTICAL INFERENCES ON ITS PARAMETERS
Abstract
A new mathematical model of discrete time series is proposed. It is called homogenous vector Markov chain of the order s with partial connections. The conditional probability distribution for this model is determined only by a few components of previous vector states. Probabilistic properties of the model are given: ergodicity conditions and conditions under which the stationary probability distribution is uniform. Consistent statistical estimators for model parameters are constructed.
About the Authors
Yu. S. Kharin
Research Institute for Applied Problems of Mathematics and Informatics of the Belarusian State University
Belarus
Corresponding Member, D. Sc. (Physics and Mathematics), Professor, Director
Belarus
Corresponding Member, D. Sc. (Physics and Mathematics), Professor, Director
M. W. Maltsew
Research Institute for Applied Problems of Mathematics and Informatics of the Belarusian State University
Belarus
Senior researcher, Head of the Research Laboratory
Belarus
Senior researcher, Head of the Research Laboratory
N. S. Sologub
Research Institute for Applied Problems of Mathematics and Informatics of the Belarusian State University
Belarus
Assistant
Belarus
Assistant
References
1. Kemeni J. G., Snell J. L. Finite Markov chains. Moscow, Nauka, 1970. 272 p. (in Russian)
2. Woterman M. S. Mathematical methods for analysis of DNA sequences. Moscow, Mir, 1999. 350 p. (in Russian)
3. Bonacich P., Liggett T. M. Asymptotics of a matrix valued Markov chain arising in sociology. Stochastic Processes and their Applications, 2003, vol. 104, is. 1, pp. 155–171. doi:10.1016/S0304-4149(02)00231-4.
4. Kharin Yu. S., Agievich S. V., Vasiliev D. V., Matveev G. V. Cryptology. Мinsk, Belarusian State University Publ., 2013. 512 p. (in Russian)
5. Zubkov A. M. Random data generators and their application. Moskovskii universitet i razvitie kriptografii v Rossii: materialy konferentsii v Moskovskom gosudarstvennom universitete [Moscow State University and the cryptography development in Russia: Proceedings of the Conference at the Moscow State University]. Moscow, Moscow Center for Continuous Mathematical Education Publ., 2003, pp. 200–206. (in Russian)
6. Dub J. L. Probability processes. Moscow, Foreign Literature Publishing House, 1956. 605 p. (in Russian)
7. Kharin Yu. Parsimonious models for high-order Markov chains and their statistical analysis. VIII World Congress on Probability and Statistics. Istanbul, Publ. House of Koc. Univ., 2012, pp. 168–169.
8. Kharin Yu. S. Markov chains with the r-partial connections and their statistical estimation. Doklady Nacional’noj akademii nauk Belarusi [Doklady of the National Academy of Sciences of Belarus], 2004, vol. 48, no. 1, pp. 40–44. (in Russian)
9. Shiryaev A. N. Probability. 2nd ed. Moscow, Nauka, 1989. 640 p. (in Russian)
10. Kharin Yu. S., Maltsev M. V. Algorithms of statistical analysis of the Markov chains with conditional memory depth. Informatika [Informatics], 2011, no. 1, pp. 34–43. (in Russian)
11. Borovkov A. A. The theory of probabilities. Moscow, Nauka, 1986. 432 p. (in Russian)
12. Basawa I. V., Prakasa Rao B. L. S. Statistical inference for stochastic processes. London, Academic Press, 1980. 438 p.
13. Kramer G. Mathematical methods of statistics. Moscow, Mir, 1975. 648 p. (in Russian)
14. Cawley G. C., Talbot N. L. C. On over-fitting in model selection and subsequent selection bias in performance evaluation. Journal of Machine Learning Research, 2010, vol. 11, pp. 2079–2107.
15. Csiszar I., Shields P. C. The consistency of the BIC Markov order estimator. The Annals of Statistics, 2000, vol. 28,no. 6, pp. 1601–1619. doi:10.1214/aos/1015957472.
Review
Views: 771
Email this article 