线性约束回归问题的几个算法

Algorithms for Linear Constrained Regression Problem

由实际问题建立的多元线性回归模型Y=Xβ+e,有时要求β满足某些线性约束条件,成为约束二次规划问题。利用数学规划方法给出求解线性约束回归问题的几个不同的算法,并给出计算实例以予比较。

Multivariate linear regression model Y=βX+ε was built from practical problems. Sometime β must satisfy some linear constrained conditions. In this paper, we present some algorithms or con-strained regression problem, applied mathematical programming methods and present computation-al examples.