Statistical inferences on transition probabilities of Markov chains under missings

The paper is devoted to hypotheses testing for finite homogeneous Markov chains under missings. The method of approximation of the likelihood function is proposed. By the proposed method, the known criteria for Markov chains, which are based on the likelihood function, may be easily transformed to criteria for Markov chains under missings. As examples, the test for a hypothetical matrix of one-step transition probabilities and the test for independence are constructed.