Detect outliers in time series data with multi-granule periodic patterns
Lin Shang · Journal of Shandong University · 2009
Contributions on outlier detection in time series data have seldom taken into account the data cyclical nature and numerical attributes values.An algorithm to find periodic patterns under different granularities was proposed,which could be used to detect outliers in time series data with numerical attributes.This method could avoid a false alarm,and experimental results showed that it could not only correctly identify multi-granule periodic patterns but also effectively detect outliers in data.Compared to outlier detection methods without periodic patterns,the results showed that it could reduce false alarms.