以黄山风景区1979-2004年的年度旅游人数数据为训练集,运用模糊聚类算法客观地分割论域,然后建立了一种模糊时间序列的加权模型以充分利用观测样本的初始信息。通过对传统的模糊时间序列模型与基于模糊聚类算法的加权模糊时间序列模型两种预测结果均方误差的比较,可见基于模糊聚类算法的加权模糊时间序列模型不但具有更高的预测精度而且也具有更好的稳定性。
Firstly ,the data of Huangshan’ s annual tourist arrivals from1979 to 2004 is taken account to the training set .T hen the method of fuzzy clustering is used to divide the discourse . In order to make full use of the initial information of observed samples ,a weighted fuzzy time series model is established .Through the mean square error’s comparison of forecasting results of two kinds of between traditional model and weighted fuzzy time series model based fuzzy clustering ,the stability of weighted fuzzy time series model is thought better .