Analytic center and maximum likelihood estimators

The purpose of this paper is to show that the analytic center, originally proposed for identification in a bounded error setting, is a maximum likelihood estimator for a class of probability distributions.