基于滑动窗口的时间序列离群数据挖掘

Outlier mining for time series based on sliding window

离群数据挖掘是数据挖掘中的重要内容。本文针对时间序列数据进行离群数据挖掘方法的研究。在引入了基于局部离群点因子的离群数据挖掘方法与时间序列上滑动窗口基础上,将二者相结合,提出了基于滑动窗口的时间序列离群数据挖掘算法,并将算法应用于海表温度数据得到海表温度的异常之处。

Outlier mining is an important part of data mining. In this paper, a research on the outlier mining method for time series data is undertaken. On the basis of the method of outlier mining based on local outlier factor and the sliding window on time series, the outlier mining algorithm for time series based on sliding window is proposed and used in sea surface temperature （SST） data to detect the anomalies about SST.