A general class of entropy statistics (Q1265618)

The author introduces a general form of entropy which is depending on four functions. We can obtain 23 different cases of entropies, known from the literature, by the choice of these functions. The article deals with the asymptotic distributions such entropies depending on some parameter which is estimated by the maximum likelihood method. The results are given in two theorems. The third part of this article is devoted to applications of the results for testing some hypotheses and to constructing confidence intervals. The article finishes by a concrete example of the Shannon entropy in the case of exponential distributions.