基于Median函数的分段回归模型及其在生物学上的应用  

Segmented Regression Model Based on Median Function and Application on Biology

作　　者：于向鸿[1] 肖阳[1] 

机构地区：[1]中国农业科学院研究生院统计室,北京100081

出　　处：《生物数学学报》2012年第3期529-536,共8页Journal of Biomathematics

基　　金：中国农业科学院基本科研业务项目(0042006004)

摘　　要：在生物学科研工作中,经常会遇到因变量和自变量之间存在着多种不同的趋势,分段回归模型可以很好的拟合变量间这种非线性趋势.本文介绍了基于Median函数的分段回归模型,可以同时对各项回归参数和转折点进行估计,最后,本文结合生物学上的实例运用SAS统计软件进行了分段回归模型的拟合.Frequently the relationship of response variable and independent variable can be seen that there are several different trends in the field of biology.Segmented regression model can fit the nonlinear trends of the variables well.In this paper,it is introduced that the segmented regression model based on median function can estimate every parameters and knots at one time. At last,at the end of paper,SAS software package has been used to fit the segmented regression model with biological examples.

关 键 词：分段回归模型 非线性回归 Median函数 SAS程序 

分 类 号：O212[理学—概率论与数理统计]