α-加权和贴近度递变模糊线性回归模型及应用

Study of α-weighted and graded closeness fuzzy linear regression model and its application

用模糊线性回归分析方法进行多元线性回归分析,能较好地处理传统模糊线性回归模型中因变量和影响变量之间关系复杂、预测结果有一定模糊性的回归分析预测问题。以因变量和各回归变量的相关系数为基础,采用α-加权和递变的贴近度标准对传统的模糊线性回归模型进行了改进,并对改进后的模型进行了评价。实际算例结果表明,改进模型的拟合精度和预测精度均有较大的提高,具有一定的实用性。

Through substitution of the fuzzy linear regression analysis for the traditional multiple linear regression analysis,the prediction issue of the regression analysis of the relationship of dependent variable and regression variable,which is complicated and with fuzzy predicting results,can be well solved.In view of the problems of the traditional fuzzy linear regression model,based on the correlation coefficient between the dependent variable and the regression variables,by using α-weighted and graded closeness standard,the traditional fuzzy linear regression model is improved and the improved model is evaluated.The actual numerical results show that the improved model is superior in model fitting and prediction accuracy,which has a certain practicality.