摘要
基于时间序列符号化后的特点,创造性地提出了一种新型相似性度量方法——符号化的统计向量空间法(SAX[1]Statistical Vector Space,SSVS)。将这种度量方法用于S&P500指数的股票数据聚类实验,并与经典相似性度量方法比较,结果表明这种新的方法能够高效地从整体趋势的角度度量时间序列的相似性,有很好的实际意义和应用前景。
Based on the feature of symbolic time series data, a new similarity measure method ( SAX Statistical Vector Space, SSVS) was cleatively brought. The stock data from the Standard & Poor 500 index was obtained, and the behavior of kinds of similarity measure methods in clustering research was compared. The experiment results showed that the new method could efficiently measure the similarity according to the whole trend by comparing with the classic methods.
出处
《计算机应用研究》
CSCD
北大核心
2007年第5期112-114,共3页
Application Research of Computers
关键词
时间序列
相似性度量
数据挖掘
符号化
time series
similarity measure
data mining
symbolic method