IRC-Safe Graph Autoencoder for Unsupervised Anomaly Detection

Atkinson, Oliver; Bhardwaj, Akanksha; Englert, Christoph; Konar, Partha; Ngairangbam, Vishal S.; Spannowsky, Michael

doi:10.3389/frai.2022.943135

IRC-Safe Graph Autoencoder for Unsupervised Anomaly Detection Thumbnail

Authors

Oliver Atkinson

Akanksha Bhardwaj

Christoph Englert

Partha Konar

Vishal S. Ngairangbam

Professor Michael Spannowsky michael.spannowsky@durham.ac.uk
Professor

Abstract

Anomaly detection through employing machine learning techniques has emerged as a novel powerful tool in the search for new physics beyond the Standard Model. Historically similar to the development of jet observables, theoretical consistency has not always assumed a central role in the fast development of algorithms and neural network architectures. In this work, we construct an infrared and collinear safe autoencoder based on graph neural networks by employing energy-weighted message passing. We demonstrate that whilst this approach has theoretically favorable properties, it also exhibits formidable sensitivity to non-QCD structures.

Citation

Atkinson, O., Bhardwaj, A., Englert, C., Konar, P., Ngairangbam, V. S., & Spannowsky, M. (2022). IRC-Safe Graph Autoencoder for Unsupervised Anomaly Detection. Frontiers in Artificial Intelligence, 5, Article 943135. https://doi.org/10.3389/frai.2022.943135

Journal Article Type	Article
Acceptance Date	Jun 23, 2022
Online Publication Date	Jul 22, 2022
Publication Date	2022
Deposit Date	Sep 5, 2022
Publicly Available Date	Sep 5, 2022
Journal	Frontiers in Artificial Intelligence
Print ISSN	2624-8212
Electronic ISSN	2624-8212
Publisher	Frontiers Media
Peer Reviewed	Peer Reviewed
Volume	5
Article Number	943135
DOI	https://doi.org/10.3389/frai.2022.943135

Files

Published Journal Article (1.2 Mb)
PDF

Publisher Licence URL
http://creativecommons.org/licenses/by/4.0/

Copyright Statement
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.