Neural Contextual Anomaly Detection for Time Series

Neural Contextual Anomaly Detection for Time Series

Chris U. Carmona, Fran├žois-Xavier Aubet, Valentin Flunkert, Jan Gasthaus

Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence
Main Track. Pages 2843-2851. https://doi.org/10.24963/ijcai.2022/394

We introduce Neural Contextual Anomaly Detection (NCAD), a framework for anomaly detection on time series that scales seamlessly from the unsupervised to supervised setting, and is applicable to both univariate and multivariate time series. This is achieved by combining recent developments in representation learning for multivariate time series, with techniques for deep anomaly detection originally developed for computer vision that we tailor to the time series setting. Our window-based approach facilitates learning the boundary between normal and anomalous classes by injecting generic synthetic anomalies into the available data. NCAD can effectively take advantage of domain knowledge and of any available training labels. We demonstrate empirically on standard benchmark datasets that our approach obtains a state-of-the-art performance in the supervised, semi-supervised, and unsupervised settings.
Keywords:
Machine Learning: Time-series; Data Streams
Data Mining: Anomaly/Outlier Detection