DETECTION METHOD AND APPARATUS FOR TIME SERIES ANOMALY POINT, AND DEVICE AND READABLE STORAGE MEDIUM

摘要

A detection method and apparatus for a time series anomaly point, and a computer device and a computer-readable storage medium, which relate to the field of artificial intelligence. The method comprises: acquiring a target window in a time series to be processed, and acquiring windows to be processed corresponding to the target window (S101), wherein the number of windows to be processed is greater than one; on the basis of a preset model, determining, according to the target window and the windows to be processed, a similarity value between the target window and each of the windows to be processed (S102); and according to the similarity values, determining whether a time point corresponding to the target window is an anomaly point (S103). By means of the method, a target window and windows to be processed in a time series are acquired, such that collected sample series have a relatively small difference; and the similarity value between the target window and each of the windows to be processed is determined by means of the preset model, thereby avoiding a relatively great error in an obtained anomalous series.