M. I. Tikhomirova, V. P. Chistyakov
On two statistics of chi-square type based on frequencies of tuples of states of a high-order Markov chain
We consider a tuple of states of an (s – 1)-order Markov chain whose transition probabilities
depend on a small part of s – 1 preceding states. We obtain limit distributions of certain
χ2-statistics X and Y based on frequencies of tuples of states of the Markov chain. For the statistic X, frequencies of tuples of only those states are used on which the transition probabilities depend, and for the statistic Y, frequencies of s-tuples without gaps. The statistical test with statistic X which distinguishes the hypotheses H1 (a high-order Markov chain) and H0 (an independent equiprobable sequence) appears to be more powerful than the test with statistic Y . The statistic Z of the Neyman–Pearson test, as well as X, depends only on frequencies of tuples with gaps. The statistics X and Y are calculated without use of distribution parameters under the hypothesis H1, and their probabilities of errors of the first and second kinds depend only on the non-centrality parameter, which is a function of transition probabilities. Thus, for these statistics the hypothesis H1 can be considered as composite.
Discrete Mathematics and Applications, Walter de Gruyter
Print ISSN: 0924-9266
Volume: 13, 07/2003
Pages: 319 - 329
Show full article (external site)
Show all available items of this journal
Search
