A method and apparatus for detecting anomalies in trading objects

Abstract

This invention provides an anomaly detection method and apparatus for trading objects, applicable to the field of computer technology. In this method, M trading elements are divided into N trading dimensions. The trading elements contained in the N trading dimensions are combined to obtain L trading objects. For each selected trading object, the anomaly detection result is determined based on the target trading data of the trading object in multiple detection periods. This invention refines different trading objects based on trading dimensions and trading elements, and obtains anomaly detection results based on the target trading data of different trading objects. Compared to anomaly detection methods that use the same set of alarm thresholds for all trading objects, this invention is more targeted, thus ensuring the accuracy of anomaly detection for trading objects with small trading volumes. Furthermore, combining the target trading data from multiple detection periods to obtain anomaly detection results further improves the accuracy of anomaly detection.

Images