偏最小二乘回归分析在均匀设计试验建模分析中的应用  

Application of Partial Least-Squares Regressive on Modeling Analysis of Uniform Design Experiment

作　　者：唐启义[1] 唐洁[1] 

机构地区：[1]浙江大学农业与生物技术学院植物保护系,杭州310029

出　　处：《数理统计与管理》2005年第5期45-49,67,共6页Journal of Applied Statistics and Management

摘　　要：本文分析了目前应用一般的最小二乘法建立均匀试验数据的二次多项式回归模型时存在的局限性,提出了应用偏最小二乘法(Partial least-square,PLS)建立二次多项式回归模型的技术,并且进一步介绍了偏最小二乘回归(PLS回归)在均匀设计中的应用。作者认为,PLS回归分析建模技术将为均匀设计的更广泛应用提供有力的技术支持。This paper analyzed the limitation of applying the least square method to establish the Second -degree Pol- ynomial Models of Uniform design data, and presented the technology of applying Partial least - squares method to establish the Second - degree Polynomial Models. Further the application of Partial least - squares regression method （PLS regression） in the Uniform Design was introduced. The author think the technolgy will provide technological support for the more extensive application of Uniform Design.

关 键 词：偏最小二乘法 均匀设计 回归分析 模型优化 

分 类 号：O212[理学—概率论与数理统计] C8[理学—数学]