TeVAE: A Variational Autoencoder Approach for Discrete Online Anomaly Detection in Variable-state Multivariate Time-series Data

📅 2024-07-09
🏛️ arXiv.org
📈 Citations: 3
Influential: 0
📄 PDF
🤖 AI Summary
Addressing the challenge of online anomaly detection in automotive testing—characterized by multivariate, state-varying time-series data, scarcity of labeled samples, stringent requirements for low false positive rate (FPR), real-time responsiveness, and root-cause interpretability—this paper proposes the Temporal Variational Autoencoder (TeVAE). TeVAE mitigates the “bypass” problem in standard VAEs via a tailored architectural design; introduces a novel sliding-window-to-continuous-time remapping mechanism that decouples anomaly localization from detection latency; and defines a new evaluation metric jointly optimizing detection delay and root-cause localization accuracy. The method is fully unsupervised and supports few-shot training. Evaluated on real-world industrial datasets, TeVAE achieves a 6% FPR and 65% detection rate—substantially outperforming established baselines—demonstrating strong engineering applicability and promising capability for root-cause analysis.

Technology Category

Application Category

📝 Abstract
As attention to recorded data grows in the realm of automotive testing and manual evaluation reaches its limits, there is a growing need for automatic online anomaly detection. This real-world data is complex in many ways and requires the modelling of testee behaviour. To address this, we propose a temporal variational autoencoder (TeVAE) that can detect anomalies with minimal false positives when trained on unlabelled data. Our approach also avoids the bypass phenomenon and introduces a new method to remap individual windows to a continuous time series. Furthermore, we propose metrics to evaluate the detection delay and root-cause capability of our approach and present results from experiments on a real-world industrial data set. When properly configured, TeVAE flags anomalies only 6% of the time wrongly and detects 65% of anomalies present. It also has the potential to perform well with a smaller training and validation subset but requires a more sophisticated threshold estimation method.
Problem

Research questions and friction points this paper is trying to address.

Detects anomalies in multivariate time-series data
Models testee behavior in automotive testing scenarios
Minimizes false positives using unlabeled training data
Innovation

Methods, ideas, or system contributions that make the work stand out.

Temporal variational autoencoder for anomaly detection
Remapping windows to continuous time series
Minimizing false positives in unlabeled data training
🔎 Similar Papers
No similar papers found.
L
Lucas Correia
Mercedes-Benz AG, Stuttgart, Germany
J
Jan-Christoph Goos
Mercedes-Benz AG, Stuttgart, Germany
P
Philipp Klein
Mercedes-Benz AG, Stuttgart, Germany
T
T. Back
Leiden University, Leiden, The Netherlands
A
Anna V. Kononova
Leiden University, Leiden, The Netherlands