On consistency of the maximum likelihood principle based tests and estimates in the case of exponential-like distributions

Rok, strany: 2001, 115 - 141

Consistency of the likelihood ratio tests is established under the assumption that the infimum of the Kullback-Leibler information number on the hypothesis of interest is a real number and certain regularity conditions are fulfilled. Under these conditions also consistency of the MLE is proved in the case that the parameter minimizing the Kullback-Leibler information number is unique. The results are proved in the multisample setting and applied to exponential distributions with unknown lower bound.

ISO 690:

Rublík, F. 2001. On consistency of the maximum likelihood principle based tests and estimates in the case of exponential-like distributions. In Tatra Mountains Mathematical Publications, vol. 22, no.2, pp. 115-141. 1210-3195.

APA:

Rublík, F. (2001). On consistency of the maximum likelihood principle based tests and estimates in the case of exponential-like distributions. Tatra Mountains Mathematical Publications, 22(2), 115-141. 1210-3195.