PLS回归在消除多元共线性中的作用

PLS regression in the function of eliminating collinearities

目的 探讨PLS回归理论在消除多元共线性中的作用。方法 利用PLS回归NIPALS算法 ,对医学实例进行分析。结果 NIPALS算法所得PLS回归模型可达到较好的拟合和预测效果。结论 当样本自变量间存在多重共线性时 。

Objective To research the NIPALS algorithm of PLS regression in the function of eliminating collinearities.Methods Medical data with NIPALS algorithm of PLS regression were analyzed. Results It showed that PLS regression model building with NIPALS algorithm got good fitting and effects of forecasting. Conclusion When collinearities exist, PLS regression is a good multiple regression.