Neural Contextual Anomaly Detection for Time Series

TL;DR

This paper presents NCAD, a scalable neural framework for anomaly detection in time series that adapts from unsupervised to supervised settings, leveraging representation learning and deep anomaly detection techniques.

Contribution

Introduces NCAD, a novel neural framework combining representation learning and deep anomaly detection tailored for diverse time series anomaly detection scenarios.

Findings

Achieves state-of-the-art results on benchmark datasets.

Effective in both unsupervised and semi-supervised settings.

Handles univariate and multivariate time series efficiently.

Abstract

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 effectively 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. Moreover, our method can effectively take advantage of all the available information, be it as domain knowledge, or as training labels in the semi-supervised setting. We demonstrate empirically on standard benchmark datasets that our approach obtains a state-of-the-art performance in these settings.