Gompertz曲线参数估计的一种加权方法  

A Weighted Method for Parameter Estimation of Gompertz Curves

作　　者：王增辉[1] 赵彦军[1] 姜淑珍[1] 黄东岩 Wang Zenghui;Zhao Yanjun;Jiang Shuzhen;Huang Dongyan(College of Humanities,Northeast Normal University,Changchun 130117,China;Key Laboratory of Bionic Engineering of Ministry of Education,Jilin University,Changchun 130025,China)

机构地区：[1]东北师范大学人文学院,长春130117 [2]吉林大学工程仿生教育部重点实验室,长春130025

出　　处：《统计与决策》2018年第20期26-29,共4页Statistics & Decision

基　　金：国家科学技术部国家级重点研发项目(2016YFD0700300)

摘　　要：对于Gompertz的参数估计，文章给出了一种加权估计方法，这种估计参数的方法是：首先用三点法或四点法来估计Gompertz曲线中的参数a，然后用加权最小二乘法估计Gompertz曲线中的参数b与k，通过实例分析表明这种加权估计参数的方法比普通的最小二乘法精度高。This paper gives a weighted estimation method for Gompertz parameter estimation. The proposed method of param- eter estimation is as follows： first, the parameters in the Gompertz curve are estimated by using the three-point or four-point meth- od [a]; and then the parameters [b] and [k] in the Gompertz curve are estimated by the weighted least square method. The example analysis shows that this weighted parameter estimation method is more accurate than the ordinary least square method.

关 键 词：Gompertz曲线 参数估计 加权最小二乘法 

分 类 号：O212[理学—概率论与数理统计]