Method and device for detecting anomaly of time sequence KPI (Key Performance Indicator) and medium
The invention discloses a time sequence KPI anomaly detection method and device and a medium, and relates to the field of cloud computing. The method comprises the following steps: acquiring time sequence data; obtaining an original significance feature sequence of the time series data; decomposing...
Saved in:
Main Author | |
---|---|
Format | Patent |
Language | Chinese English |
Published |
02.08.2022
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention discloses a time sequence KPI anomaly detection method and device and a medium, and relates to the field of cloud computing. The method comprises the following steps: acquiring time sequence data; obtaining an original significance feature sequence of the time series data; decomposing the original saliency feature sequence and obtaining each component after decomposition; fitting the components and adding the values on the fitted components so as to obtain a reconstructed significance feature sequence; obtaining a difference value between the original significance feature sequence and the reconstructed significance feature sequence; and obtaining abnormal data in the time series data according to the difference value. Compared with a method for determining the abnormal point in the time series data by setting a threshold value according to experience, the method provided by the invention is based on the time series data, the saliency characteristics are extracted from the time series data, the e |
---|---|
Bibliography: | Application Number: CN202210467233 |