A Discretized Approach to Flexibly Fit Generalized Lambda Distributions to Data

被引：11
|
作者：
Su, Steve [1 ,2 ]
机构：
[1] George Inst Int Hlth, Epistat Div, Sydney, NSW, Australia
[2] Univ Sydney, George Inst, Epistat Div, Sydney, NSW, Australia
来源：
JOURNAL OF MODERN APPLIED STATISTICAL METHODS | 2005年 / 4卷 / 02期
关键词：
generalized lambda distributions; quantile distributions; fitting distributions to data;
D O I：
10.22237/jmasm/1130803560
中图分类号：
O21 [概率论与数理统计]; C8 [统计学];
学科分类号：
020208 ; 070103 ; 0714 ;
摘要：
This article presents a flexible approach to fit statistical distribution to data. It optimizes the bin-width of data histogram to find a suitable generalized lambda distribution. In addition to the default optimization, this approach provides additional flexibility akin to the concepts of loess and kernel smoothing, which allow the users to determine the amount of details they would like to smooth over the data. The approach presented in this article will allow users to visually compare and choose the parameters of generalized lambda distribution that best suit their purposes of study.
引用
收藏
页码：408 / 424
页数：17
相关论文
共 50 条
12345