MODELIZATION IN R-TH ORDER MARKOV CHAINS
WITH DEPENDENCY ON COVARIATES
Manuel MOLINA-FERNANDEZ1
Miguel GONZALEZ-VELASCO[1]
§
ABSTRACT:
In this work, the multiple-group logistic regression model is used by the
modelization of the transition probabilities in r-th order Markov chains
depending on covariates. The maximum likelihood estimation of the parameters is
considered and a method to fit a more appropriate model is proposed. Some
extensions of the basic model are indicated.
§
KEYWORDS:
Finite Markov chain; multiple-group logistic regression model; maximum
likelihood inference; exponential family.