Intelligent Forecast And Decision-Making Based On Vague Set Thoery

This paper mainly studied the problems of forecasting and decision making. In the forecasting problem, we constructed vague time series to forecast problem based on the time series and vague sets theory. In the decision making problem, we presented a new evaluation method of vague sets for vague multi-attribute decision making. Main works are as follows:(1)Let vague sets theory combine with time series and made vague sets instead of the point of time series. We considered adequately the hesitated degree of vague sets and imported it into the time series, and then constructed a forecasting model based on vague time series, through the case analysis to illustrate the feasibility of the model. Meanwhile, in the view of the historical data are large, presented an improved vague time series model.(2)According to the characteristics that vague time series could not forecast dynamically, in the basis of vague time series and sliding window, we constructed a forecasting model based on time-variant vague time series, through the case to analyzed and compared the effectiveness of the method.(3)According to the characteristics that vague sets theory could express true-membership、false-membership and hesitating-membership, we took these memberships map in the triangular rectangular coordinates system. Through the distance of the point in the coordinates system between vague set and the optimization to evaluate vague set, we could judge the good or bad of vague sets. Meanwhile, the distance method needs to keep projection comparison, so we improved it without projection comparison. Finally, we made this method expand into vague set multi-attribute decision making problem.