Online Anomaly Detection in Multivariate Settings

Mahsa Mozaffari, Yasin Yilmaz

Abstract: This paper considers the real-time and nonparametric detection of anomalies in high-dimensional systems. The goal is to detect anomalies quickly and accurately such that the appropriate countermeasures could be taken before any possible harm is caused by the anomalous event. We propose a k NN-based sequential anomaly detection method in both semi-supervised and supervised settings. We prove that the proposed method is asymptotically optimum in the minimax sense under certain conditions in terms of minimizing the average detection delay for a given false alarm constraint. The proposed method is shown to be capable of multivariate anomaly detection and also scalable to high-dimensional datasets. We further propose an online learning scheme that combines the desirable properties of our semi-supervised and supervised methods.