Kybernetika 33 no. 2, 133-159, 1997

Asymptotic results in parameter estimation for Gibbs random fields

Martin Janžura


Both the maximum likelihood estimate and a class of the maximum pseudo-likelihood estimates for parameters of Gibbs random fields are introduced, and their asymptotic properties, namely the consistency, the asymptotic normality, and the asymptotic efficiency, are studied, as well as the interrelations between the particular estimators and their respective properties.


62M40, 62F12, 93E10