摘要
本文采用偏最小二乘回归模型(PLS),以泰国菠萝贸易为例,通过变量投影重要性准则筛选自变量,由交叉有效性提取主成分,进而建立偏最小二乘回归模型。深入分析了各指标对泰国菠萝出口贸易的影响。研究表明泰国菠萝出口与原料价格及工厂生产加工速度密切相关,并且偏最小二乘回归的拟合效果优于普通最小二乘回归。
Based on the partial least (PLS) method, for example of the factors that effect on Thai pineapple exports, we select variables by variable importance in projection and extract principal components via cross validation, then establish the partial least square models. We analyze the impact of each indicator on Thai pineapple exports deeply. Studies have shown that Thai pineapple exports are closed related to the raw material prices and the speed of production and processing in factories, and the effects of partial least squares regression fit better than ordinary least squares regression.
出处
《价值工程》
2014年第31期9-11,共3页
Value Engineering
基金
国家自然科学基金(11201200
11261025)
云南省自然科学基金(KKSY201207004)
关键词
泰国菠萝
偏最小二乘回归
交叉有效性
Thai pineapple
partial least squares regression
cross validation