Summary of Tevae: a Variational Autoencoder Approach For Discrete Online Anomaly Detection in Variable-state Multivariate Time-series Data, by Lucas Correia et al.
TeVAE: A Variational Autoencoder Approach for Discrete Online Anomaly Detection in Variable-state Multivariate Time-series Data
by Lucas Correia, Jan-Christoph Goos, Philipp Klein, Thomas Bäck, Anna V. Kononova
First submitted to arxiv on: 9 Jul 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed temporal variational autoencoder (TeVAE) is a novel approach for automatic online anomaly detection in automotive testing, aiming to minimize false positives when trained on unlabelled data. This real-world problem requires modeling testee behavior and handling complex data. TeVAE avoids the bypass phenomenon by remapping individual windows to a continuous time series. The proposed method also introduces metrics to evaluate detection delay and root-cause capability. Experiments on industrial data sets show that when properly configured, TeVAE flags anomalies only 6% of the time wrongly and detects 65% of present anomalies. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary TeVAE is a new way to find problems in car tests without looking at all the data. This helps by understanding how people behave during these tests. The model is good at not finding fake problems and finds most real issues. It can even work well with only a little training data, but needs better ways to decide when to alert. |
Keywords
» Artificial intelligence » Anomaly detection » Time series » Variational autoencoder
