基于偏最小二乘回归方法的产品满意度回归模型研究  

Research on Regression Model of Product Customer Satisfaction Based on PLS

作　　者：华尔天[1] 毛明杰[2] 叶飞帆[3] 裴仁清[1] 

机构地区：[1]上海大学,上海200072 [2]浙江工业大学,杭州310023 [3]宁波大学,宁波315211

出　　处：《中国机械工程》2005年第3期192-195,共4页China Mechanical Engineering

基　　金：浙江省自然科学基金资助项目(502156)

摘　　要：　　将产品满意度与主要产品性价要素直接联系起来,基于偏最小二乘回归方法建立了产品满意度回归模型,较好地解决了由变量多重相关性等引起的满意度估计量数据失真问题;不断地通过主成分提取和还原计算,相对准确地确定了各产品性价要素在产品满意度计算中的权重,从而为产品改型设计提供了重要依据。Product customer satisfaction (PCS) was directly connected with the key elements of product price and performance, and the regression model of PCS was established based on partial least-square(PLS) regression model, so that the harmful effects due to multicollinearity were corrected, and regression parameters of the model can be properly confirmed by extracting principal components. To do so, it has offered the important basis for the issues that the products are innovated and designed.

关 键 词：偏最小二乘回归方法 产品满意度 交叉有效性 要素权重 

分 类 号：F406.3[经济管理—产业经济]