基于距离和密度的时间序列异常检测方法研究

Research on discords detect on time series based on distance and density

在时间序列的GMBR表示的基础上,首次提出将基于距离和基于密度的时间序列检测方法结合,给出了时间序列模式异常的定义,并用"异常特征值"来衡量时间序列模式的异常程度。根据所提出的模式异常的定义,在强力搜索算法的基础之上提出了新的时间序列异常检测算法GMBR-DD(Grid Minimum Bounding Rectangle-Discords Detect),该算法将基于距离和基于密度的异常检测方法结合,能够高效地发现时间序列中的异常模式。通过三组实验数据,对提出的异常时间序列定义和时间序列的异常检测算法进行了验证,实验结果表明所提出的时间序列异常检测算法能够有效地发现时间序列的异常变动,为决策提供了很好的平台和有力的工具。

It proposes the definition of the discords detect of time series based on the representation of the GMBR （Grid Minimum Bounding Rectangle） and it is the first time to combine the distance measure method with density. It uses the ＂detect eigenvalue＂ to weigh the detect degree of the time series. Based on the proposed definition of the discords detect, it gives the new discords detect algorithm named GMBR-DD （Grid Minimum Bounding RectangleDiscords Detect）. This algorithm can find the discords time series with high-effect. It validates the definition and the proposed algorithm through three groups of the data. The experimental results show that the algorithm can catch the discords time series and the definition is reasonable. So the production provides a very effect fiat roof and a powerful tool in data mining of time series.