Estimation In The Location-scale-shape Model Of Type I Generalized Logistic Distribution

This thesis consists of three parts: the first part is about the basic properties of generalized logistic distribution(GLD) and the research status of its parameter estimation; the second part provides parameter estimations of GLD using a variety of estimation methods and the large sample properties of some estimations; the third part is simulation study and we compare all the results in this part.This thesis provides the method of L-moment, probability weighted moment estimator,generalized probability weighted moment estimator, maximum likelihood estimator and least square estimator in the location-scale-shape model of type I GLD. This thesis also provides some large sample properties of the estimator, and compare the performances of these procedures through an extensive numerical simulation. Least square estimator and generalized probability weighted moment estimator of three-parameter type I GLD are put forward for the first time. After the simulation, we note that least square estimator is stable, but within no one parameter range is least square estimation absolutely superior to other methods; generalized probability weighted moment estimator is probably best used to estimate the shape parameters,but does not show certain regularity. They can all be further optimized. This thesis also provides an idea that we can use the basic feature and the scope of application of several methods to do mixed estimation.