摘要
在社会和经济等领域中存在大量的异方差数据,而且人们非常关注方差的变化,所以在研究社会经济现象时方差建模与均值建模同等重要;相对于对称分布,偏态分布更能获得更全面准确、更及时有效的信息,文章基于以上两点,研究提出基于偏t正态分布(StN)的联合位置与尺度模型,并给出该模型参数的极大似然估计,模拟和实例研究结果表明该模型和方法是有用和有效的.
In social and economic fields,there are a lot of heteroscedastic data,and we are very concerned about the change of the variance, thus modeling of the variance can be as im- portant as that of the mean in the social economic phenomenon. On the other hand, compared with the symmetrical distribution, skewness distribution can obtain more accurate and more timely information. Based on the above two points, we propose a joint location and scale mod- els based on the skew- t -normal distribution,and investigate the maximum likelihood estima- tor of this model. Simulation studies and a real example show that this model and method are useful and effective.
出处
《应用数学》
CSCD
北大核心
2013年第3期671-676,共6页
Mathematica Applicata
基金
国家自然科学基金资助项目(11261025
11126309)
云南省自然科学基金资助项目(2009ZC039M
2011FB016
2011FZ044)
关键词
StN分布
联合位置与尺度模型
极大似然估计
Skew- t-normal distribution~ Joint location and scale model~ Maximum Hkeli hood estimation
