摘要
文章针对自变量的系数和输出因变量是模糊数的模糊线性回归模型,引入模糊数的可能性均值和方差,构建模糊数的可能性均值-方差距离测度。借助于最小一乘法原理,给出了使得可能性均值-方差距离误差达到最小的回归系数估计值的线性规划模型。最后给出实例说明了方法的有效性和可行性。
Aiming at the fuzzy linear regression model whose coefficient of independent variable and output dependent variable are fuzzy numbers, this paper constructs fuzzy number's possibilistic mean-variance distance measurement. With the aid of the least absolute principle, the paper presents a linear programming model for the estimation of the regression coefficient with the minimum possibilistic mean-variance distance error. Finally the paper uses two examples of symmetric and asymmetric triangular fuzzy numbers to illustrate the effectiveness and feasibility of the proposed method.
引文
[1]Tanaka H,Uejima S,Asai K.Linear Regression Analysis With FuzzyModel[J].IEEE Transactions on Systems Man,and Cybernetics,1982,(12).
[2]任燕,郭嗣琮.基于结构元最小二乘序的模糊线性回归[J].模糊系统与数学,2015,29(1).
[3]李俊红,曾文艺.基于梯形模糊数的模糊最小一乘回归模型[J].系统工程理论与实践,2015,35(6).
[4]彭宇文,郭莉莎,毛超.基于改进模拟退火算法的模糊回归参数估计[J].统计与决策,2014,(1).
[5]岳立柱.系数为一般模糊数的多元线性回归模型[J].统计与决策,2015,(3).
[6]Carlsson C,Fullér R.On Possibilistic Mean Valueand Variance of Fuzzy Numbers[J].Fuzzy Sets andSystems,2001,(122).
[7]Nasrabad M M.Narsabad E.A Mathematical-progrm-ming Approacg to Fuzzy Linear Regression Analysis[J].Applied Mathematics and Computation,2004,(155).
[8]Coppia R,Urso P D,Giordani P,et al.Least SquaresEstimation of a Linear Regression Model With LRFuzzy Response[J].Computational Statistics&DataAnalysis,2006,(51).
[9]黄华,易正俊.修正的模糊最小一乘线性回归[J].模糊系统与数学,2012,26(1).