基于时态边缘算子的时间序列分段线性表示  

Piecewise linear representation of time series based on temporal edge operator

作　　者：肖辉[1] 马海兵[1] 龚薇[1] 

机构地区：[1]浙江大学宁波理工学院,中国人民解放军92664部队,宁波大学商学院,浙江宁波315100,山东青岛266032,浙江宁波315211

出　　处：《计算机工程与应用》2008年第19期156-159,共4页Computer Engineering and Applications

基　　金：宁波市自然科学基金(the Natural Science Foundation of Ningbo City of China under Grant No.2007A610043)

摘　　要：时间序列的分段线性表示算法通常基于单一的启发式规则,难以适用于不同数据特征的时间序列。借鉴了边缘算子的思想来提取时间序列的边缘点,提出了一种基于时态边缘算子的时间序列分段线性表示算法。在来自不同领域的公开数据集上进行的实验结果表明:与两种主要的分段线性表示算法相比,该算法具有更好的拟合性能,并且更为稳定,能够适用于各类不同数据特征的时间序列。Piecewise linear representation algorithms of time series are usually based on a single heuristic rules,difficult to apply in different characteristics of time series data.Drawing on the idea of edge operator to extract time series edge points,propose a novel piecewise linear representation algorithm that based on temporal edge operator.The experiments on public datasets from several different fields show that compared with two main piecewise linear representation algorithms,the proposed algorithm appears better fitting error and high stability which can be applied to different characteristics of time series data.

关 键 词：时间序列 分段线性表示 拟合误差 时态边缘算子 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]