摘要
对基于最小二乘准则的回归方法进行了分析,指出该方法容易受到野点信号的干扰.最小一乘准则虽然克服了最小二乘准则的不足,但如何在其基础上建立与实现回归算法一直是一个难题.论文中提出利用线性规划的方法建立该算法,并采用Matlab语言予以实现.
The regression method based on least-square criteria is analyzed which is easy to be disturbed by noise data. The least-absolute criteria can overcome the drawbacks of least-square criteria, butjhow to design and implement a regression algorithm based on it is a difficult problem. In this paper, a least-absolute regression algorithm using linear programming is proposed and realized by matlab.
出处
《五邑大学学报（自然科学版）》
CAS
2006年第1期5-8,21,共5页
Journal of Wuyi University(Natural Science Edition)
基金
广东省自然科学基金资助项目(032353)
