摘要
分片逆回归是最近提出的一种多元数据分析方法.这是一种有效的降维方法.使用该方法的关键点在于能给出条件协差阵一个较好的估计.为此目的,本文基于拟残差给出了一个估计,并且研究了它的渐近性质,最后给出了部分模拟结果.
Li (1989,1991) and Duan and Li (1991) proposed a new technique for analysing the multivariate data: sliced inverse regression. The key step to employing the new technique is the requirement of an estimate of covariance matrix. In this paper, a Quasi-residuals-based estimate is given and asymptotical properties is obtained. Sample properties are investigated in a simulation study.
出处
《系统科学与数学》
CSCD
北大核心
2005年第3期348-355,共8页
Journal of Systems Science and Mathematical Sciences
