The Reduction Algorithm Of Fuzzy Time Series Model Based On The Charachristic Expansion Method

The study of fuzzy time series has increasingly attracted much attention due to its salient capabilities of tackling uncertainty and vagueness inherent in data collected. In order to get higher precision, a new data fuzzification method is presented by a kind of new fuzzy sets based on distance measure. Although this method has certain advantages, it also has some shortcomings. The fuzzy inference model with redundant fuzzy rules will result in increasing computing complexity and decreasing accuracy. Hence the reduction algorithm of fuzzy rules is proposed using of the characteristic expansion method and discussed in the case of the equal and the unequal interval differentiate.In the universe of discourse partition with equal interval, a new data fuzzification method is presented by a kind of new fuzzy sets based on distance measure and the fuzzy rules are established. Then, the characteristic coefficients between the minor premises and the fuzzy inference rules are calculated by the characteristic expansion method. Fuzzy rules are reducted and optimized by setting the threshold value. Finally, defuzzify the forecasting fuzzy sets using of the weighted average method. Through the forecasting of Alabama university enrollments, results show that the proposed reduction algorithm is effective.In the universe of discourse partition with unequal interval, the FCM algorithm is used to divide the universe of discourse, and combined with the proposed reduction algorithm in order to further optimize the fuzzy time series model. Through the forecasting of Alabama university enrollments, results further show that the proposed reduction algorithm is effective.