Виявлення колективних аномалій у часових рядах

Abstract

Anomaly detection for time series data has been an important area of research for a long time. The main work on anomaly detection techniques has focused on statistical approaches. In recent years, more and more machine learning algorithms have been developed to detect anomalies in time series. There are three basic types of anomalies, this work we devoted to the study of collective anomalies as the most complex and frequently encountered type of anomalies. The implementation of this type of anomalies was carried out by the k-means method as the most universal method in our opinion.