Generalizations of Tukey-Lambda Distributions
Czesław Domański, Katarzyna Bolonek-Lasoń · University of Lodz Repository (University of Łódź) · 2013
The Generalized Lambda Distribution (GλD) is a four-parameter generalization of Tukey’s Lambda family. Several methods for estimating the parameters of the GλD have been reported in the literature, but the most popular is the moment-method matching proposed by Ramberg and Schmeiser (1974). One criticism of the GλD referred to above is that the shape parameters also determine skewness. It seems reasonable that there should be three linear parameters determining position, scale, and skewness and two parameters determining the shapes of the two tails. This suggests a natural generalization of the GλD to give a five-parameter lambda distribution (FPLD). The aim of paper is to show that the GλD and the FPLD describe empirical distribution quite well.