Using maximum weighted likelihood to derive Lehmer and Hölder mean families

In this paper, we establish the links between the Lehmer and Hölder mean families and maximum weighted likelihood estimator. Considering the regular one-parameter exponential family of probability density functions, we show that the maximum weighted likelihood of the parameter is a generalized weighted mean family from which Lehmer and Hölder mean families are derived. Some of the outcomes obtained provide a probabilistic interpretation of these mean families and could therefore broaden their uses in various applications.