首页 | 本学科首页   官方微博 | 高级检索  
     

系数为LR-型模糊数的模糊线性最小二乘回归
引用本文:梁艳,魏立力. 系数为LR-型模糊数的模糊线性最小二乘回归[J]. 模糊系统与数学, 2007, 21(3): 111-117
作者姓名:梁艳  魏立力
作者单位:南京航空航天大学,金城学院,江苏,南京,210016;宁夏大学,数学计算机学院,宁夏,银川,750021;宁夏大学,数学计算机学院,宁夏,银川,750021
基金项目:教育部科学技术研究重大项目(206159),宁夏自然科学基金资助项目(NZ0516)
摘    要:针对输入、输出以及系数为LR-型模糊数的情况,建立模糊线性回归模型,提出该模型的最小二乘估计以及模型性能评价方法。当输入、输出以及系数都退化为精确值时,该估计退化为经典的最小二乘估计。该方法不仅适用于三角模糊数,也适用于其它LR-型模糊数(如指数型模糊数)。数值模拟表明,该方法的拟合效果较好。

关 键 词:LR-型模糊数  模糊线性回归  最小二乘估计
文章编号:1001-7402(2007)03-0111-07
收稿时间:2006-04-01
修稿时间:2006-04-01

Fuzzy Linear Least-squares Regression with LR-type Fuzzy Coefficients
LIANG Yan,WEI Li-li. Fuzzy Linear Least-squares Regression with LR-type Fuzzy Coefficients[J]. Fuzzy Systems and Mathematics, 2007, 21(3): 111-117
Authors:LIANG Yan  WEI Li-li
Affiliation:1. College of JingCheng,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China; 2. School of Mathmatics and Computer Science,Ningxia University,Yinchuan 750021,China
Abstract:In this paper,a fuzzy linear regression model is conducted,in which all of the inputs,outputs and coefficients are LR-type fuzzy numbers.The least-squares estimation and the performance evaluation method of the model are proposed.Especially,when the inputs,outputs and coefficients are crisp data,the estimation will be classical least squares estimation.The method is not only fits for the trian gular fuzzy numbers used commonly,but also fits for other LR-type fuzzy number,such as exponential-type fuzzy numbers.The experimental results show that this method have good fitness.
Keywords:LR-type Fuzzy Number  Fuzzy Linear Regression  Least-squares Estimation
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号