A new forecasting method of fuzzy time series model
QU Hong-wei · Kongzhi yu juece · 2013
In view of the problems that the existing fuzzy time series forecasting methods lack persuasiveness in partitioning interval and data fuzzifying.Therefore firstly,a formula contained distance parameter is proposed to calculate the cluster number,and then based on fuzzy-means(FCM) and parameters’ adjusting,unequal-sized intervals are obtained.Then a definition method of the fuzzy sets is objectively given by distance in data fuzzification.Meanwhile,the optimal forecasting results and cluster number are determined by the smallest standard error(RMSE).Finally,the forecasting of Alabama university enrollments shows the effectiveness of the proposed method.