Application of a Fuzzy method for predicting based on high-order time series

In this paper, we propose a new fuzzy prediction novel based on the higher order fuzzy time series. The proposed model is based on the higher order fuzzy time series prediction computation approach which renders a better performance in order to solve the problems of higher order fuzzy time series. The performance of the approach is represented so that after the fuzzification of time series and creating the logical fuzzy relations, some specific computations are calculated and a set of features are gained, using the lower limit of the predicting element’s range and its consecutive range, and also the resulted difference of sequential elements. In order to choose the right feature among the set, we define a term so that the features should be involved in the predicting element’s range and after defining some functions in order to calculate the membership degree of each feature, the qualified features are multiplied by their membership degree and lastly the median of the predicting element’s range is added to their sum and then divided by their sum of membership degree plus one. The yielded score is the predicted crisp value of considered element. In order to decide the precision of the prediction’s rate, we compare the proposed model to other methods using the mean square error and the average error. This method is implemented on the Alabama University’s enrollment database and less error is found in comparison to the other methods.