Generalized skew normal model

被引:55
作者
Gupta, RC [1 ]
Gupta, RD
机构
[1] Univ Maine, Dept Math & Stat, Orono, ME 04469 USA
[2] Univ New Brunswick, Dept Comp Sci & Appl Stat, Fredericton, NB E3B 5A3, Canada
关键词
normal distribution; skewness; log concave; stochastic ordering; maximum likelihood estimator;
D O I
10.1007/BF02595784
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
The skew normal distribution proposed by Azzalini (1985) can be a suitable model for the analysis of data exhibiting a unimodal density function having some skewness present., a structure often occurring in data analysis. In this paper, we study a generalization of the basic Azzalini model proposed by Balakrishnan, as a discussant of Arnold and Beaver (2002). The basic structural properties of the model including the reliability properties are presented. Estimation and testing of hypothesis of the skew parameter are discussed. Some comparisons of the models in terms of mean, variance and skewness are provided. Two data sets are analyzed.
引用
收藏
页码:501 / 524
页数:24
相关论文
共 50 条