摘要
本文再次指出:最小平方距离法(LSD),也可叫最小模方法,是从解决多维空间n个点的超平面拟合问题而提取的;通过对p个随机变量的n组观测值,此方法是探求它们之间是否存在隐式线形函数关系的好方法;使用此法,可以得出隐式线形函数关系较好的参数估计。
In this paper point: The method of least square distance (LSD) based on the idea that N points of multi-way space close in super-plane. Upon sample of multi-way variance, the method of LSD is good method of the preferences of recessive linear functions, by the method of LSD may be obtain fairly good estimates of parameter of recessive linear functions.
出处
《数理统计与管理》
CSSCI
北大核心
2013年第3期468-475,共8页
Journal of Applied Statistics and Management
关键词
超平面拟合
最小平方距离法(LSD)
隐式线性函数
参数估计
close in super-plane, the method of least square distances (LSD), the recessive linear functions, the estimates of parameter