Weighted fuzzy linear regression prediction

摘要

In order to solve the problems of dynamic forecasting with fuzzy information, the author puts forward a improved fuzzy linear regression forecasting method, weighted fuzzy linear regression forecasting method, according to the important degree of regression variables to determine the weight of each term in the objective function and the historical data to determine the important degree of similarity criteria, while the model identification in fuzzy coefficient, fuzzy coefficient of symmetrical triangular fuzzy numbers, this gives the description of symmetric triangular fuzzy number approximation degree of the fitting degree of definition and expression it gives the symmetric fuzzy number fuzzy degree is defined and its expression. The example shows that the model has high precision.