基于正交函数系和FCM算法,提出了一种新的时间序列聚类的方法.该方法首先通过一个非线性映射,将长度为$n$的时间序列映射到$L_2$空间, 然后通过计算函数之间的距离得到时间序列之间的相似度.在此基础上,经过FCM 算法实现时间序列的聚类.该方法克服了时间序列的高维数特征为时间序列聚类带来的计算困难.实验结果表明,对高维的时间序列,该方法在压缩率达到$80\mathrm{\%}$ 的情况下,依然具有良好的聚类效果.

Based on the orthogonal function system and FCM algorithm, this paper proposed a method to clustering time series. The time series with length n are first mapped in $L_2$ space. Then the similarity of series will be obtained by computing the distance among functions. Last, the cluster results of time series have been achieved by FCM algorithm according to these similarities. The advantage of the proposal is to greatly reduce the amount of calculation caused by clustering the high dimensional time series. The experimental results show that the desired cluster results can also be obtained under the case of $80\mathrm{\%}$ compressibility.