TOPSIS Method And Its Application To Multiple Attribute Decision-Making And Combination Forecast

TOPSIS method,which has been widely used in economy,finance,management,military,is a common method of multiple attributes decision-making.Currently,there are two problems in the study of measuring the difference of two triangular fuzzy numbers.For one thing,the importance of each endpoint is equally treated.For another,the former papers take little consideration into the cross impact between the endpoints.Though the information aggregating operators of combination forecast are widely used in the decision-making model,a few decision-making ideas are applied in the combination forecast.In fact,the question of the determination of the most optimal weight coefficient of combination forecast actual is a decision-making problem under some fixed attributes.Based on the above ideas,TOPSIS method and its application to multiple attributes decision-making and combination forecast is studied in this dissertation as follows:(1)This dissertation,in which the difference of each endpoint and the cross impact between the three points of triangular fuzzy numbers is mentioned,is motivated by the idea of definite integral.To regard the membership function as the integrand,a novel distance function is proposed.And it is pointed out that it actual is a kind of distance between triangular fuzzy numbers.This dissertation applies this distance and fuzzy TOPSIS method to multiple attribute decision making.A practical example-is illustrated to show-that the method has effectiveness and simple calculation.(2)The combination forecasts,which are composed of single forecast with different weights,are viewed as alternatives in this dissertation.And the mean and variance of the accuracy of the forecasts are looked upon as attributes.Then the problem how to allocate the weight coefficient of combination forecast model is converted to a decision-making problem.A combination forecast model with multiple targets is presented based on approach degree.A stock example is illustrated to show that the method has effectiveness.(3)As the forecasted information complicated,the combination forecast models with real numbers do not apply in this case.This dissertation generalizes the forecasted information to triangular fuzzy numbers.What is more,at each time point,the square sum of error of three endpoints are intended as attributes.Then A decision-making model is proposed at each time based on TOPSIS method.The close degrees are served as the induced value of the single forecast at every time point.A fuzzy combination forecast model is presented based on the IOWA operator and fuzzy TOPSIS method.A stock example is illustrated to show that the method has effectiveness.