DATA ANOMALY DETECTION METHOD AND APPARATUS, AND ELECTRONIC DEVICE AND STORAGE MEDIUM

摘要

A data anomaly detection method and apparatus, and an electronic device and a computer-readable storage medium. The method comprises: acquiring a standard training data set that includes missing data and an anomaly detection model framework that comprises a variational lower limit function; adjusting the variational lower limit function by using the missing data so as to obtain an optimized variational lower limit function; training the anomaly detection model framework by using the standard training data set so as to obtain an anomaly detection model; performing, by using the anomaly detection model, detection on a data set to be subjected to detection and obtaining a reconstruction probability of data to be tested; and if there is target data to be tested with a reconstruction probability greater than or equal to a reconstruction threshold value, determining that the target data to be tested is abnormal data. By means of the method, the efficiency and accuracy of key performance indicator (KPI) anomaly detection can be improved.