Analysis of Partial Least Squares Modeling and Multi-collinearity Ability

作者：Xia, Li^*

来源：Agro Food Industry Hi-Tech, 2017, 28(1): 885-889.

• Partial least squares
• Multi-collinearity
• Inhibition ability
• Application range

Partial least squares regression method is also considered as the second generation regression method, which has strong processing ability on multi-collinearity of linear regression. Based on partial least squares method, the existing problems were analyzed. Then the four-step modeling process of partial least squares was introduced from the perspective of mathematics. Ultimately, the inhibition ability of least square method in the problem of multi-collinearity was determined by means of mathematical induction method. Due to the limitation of partial least squares, this paper made some improvements and found that the stability of the improved partial least squares method is better than that of the traditional partial least squares method. Combined with specific theories and related examples, water supply capacity of the partial least squares method has been predicted and analyzed.

• 出版日期2017-2
• 单位武汉科技大学