The Interval Combination Forecasting Based On The Approach Degree And Distance Measure

As for the combination forecasting with the interval values, we establish three new combination forecasting models, and study the basic property of these models. The main results are as follows.(1) The Euclidean distance of the interval valued fuzzy sets are introduced during chapter three. A new model is established with the criterion of minimizing the Euclidean distance between the forecast value and the actual value. We define the dominance and the redundant measure of the model. The existence of superior combination forecasting and the redundant information are also discussed.(2) The biggest-smallest approach degree are used to propose a new combination forecasting model with the criterion of minimizing both the biggest-smallest approach degree of lower limit and the upper limit series. Concepts of the dominance and the redundant measure of the model are given. Based on these concepts, we study the existence of superior combination forecasting and the redundant information.(3) A new variant weights combination forecasting model is proposed in the last chapter. We construct the combination forecasting model based on maximum-minimum closeness degree and induced ordered weighted averaging operator..