基于Logistic回归的社区满意度模型_耿金花.pdfVIP

DOI ：10.13195/j.cd.2007.11.107.gengjh.021 第22 卷第11 期 控 制 与 决 策 2007 年 11 月 Vol. 22 No . 11 Control and Decision Nov. 2007 : 1001-0920(2007) 11-1305-04 Logistic 1 1 1 1, 2 耿金花, 高齐圣 , 任敬喜 , 张嗣瀛 ( 1. , 266071; 2. , 110004) : 在论述顾客满意度模型是具有二分类因变量的非线性模型的基础上, 出一种基于Logistic 回归的联立方 程模型. 首先采用因子分析法 取影响满意度和忠诚度的潜在变量; 然后对它们满意与否忠诚与否进行 Logistic 回 归, 建立 一个递归联立方程模型; 最后结合社区满意度实例进行了实证研究. : Logistic 回归; 联立方程; 因子分析; 社区满意度 : N945 : A Community satisfaction model based on Logistic regression 1 1 1 1, 2 GEN GJ in-hua , GAO Qi-sheng , REN J ing-xi , ZH A NGSi-ying ( 1. Com lexity Science Institute, Qingdao University , Qingdao 266071, China; 2. College of Information Science and Engineering, Northeastern University , Shenyang 110004, China. Corres ondent: GAO Qi-sheng, E-mail: gaoqisheng@ qdu. edu. cn) Abstract: Based on the issue that customer satisfaction model is a nonlinear one w ith dichotomous de endent variable, a kind of Logistic regression simultaneous equations model is ro osed. Latent factors influencing satisfaction and loyalty are extracted by factor analysis. Logistic regression analysis for the satisfaction and loyalty of theses latent factors is done, and a recursive simultaneous equation is established. Finally , an exam le about community satisfaction validates the model. Key words: Logistic regression; Simultaneous equations; Factor analysis; Community satisfaction 1 引 言 , [3-6] / , [7-10] 0 , . . Cardozo 1965