摘要
首先,提出了基于Kmeans算法的非等分论域划分方法.其次,针对传统数据模糊化存在的不足,对数据模糊化方法进行了改进.最后,将模型应用于对上海市消费价格总指数的预测,并通过与现有方法进行对比,验证了模型的有效性.
First of all,this paper puts forward the theory of unequal-sized intervals partitioning based on Kmeans algorithm.Secondly,aiming at the shortage of the traditional data fuzzification,fuzzy method is improved on data.Finally,the model is applied in predicting the total consumer price index in Shanghai,and compared with existing methods,verify the validity of the model.
出处
《应用泛函分析学报》
2015年第1期58-65,共8页
Acta Analysis Functionalis Applicata
关键词
模糊时间序列
非等分论域划分
数据模糊化
fuzzy time series
unequal-sized intervals partitioning
fuzzy data
