Using a nested anomaly detection machine learning algorithm to study the neutral triple gauge couplings at an \texorpdfstring{$e^+e^-$}{e+e-} collider

TL;DR

This paper introduces a nested anomaly detection algorithm tailored for studying neutral triple gauge couplings at future electron-positron colliders, addressing limitations of traditional anomaly detection in the presence of interference effects.

Contribution

The paper proposes a novel nested anomaly detection algorithm that improves physics analysis when standard anomaly detection assumptions do not hold, especially for interference scenarios.

Findings

Effective in studying neutral triple gauge couplings at CEPC, ILC, FCC-ee

Outperforms traditional anomaly detection in interference scenarios

Applicable as a complementary method to existing anomaly detection algorithms

Abstract

Anomaly detection algorithms have been proved to be useful in the search of new physics beyond the Standard Model. However, a prerequisite for using an anomaly detection algorithm is that the signal to be sought is indeed anomalous. This does not always hold true, for example when interference between new physics and the Standard Model becomes important. In this case, the search of new physics is no longer an anomaly detection. To overcome this difficulty, we propose a nested anomaly detection algorithm, which appears to be useful in the study of neutral triple gauge couplings at the CEPC, the ILC and the FCC-ee. Our approach inherits the advantages of the anomaly detection algorithm been nested, while at the same time, it is no longer an anomaly detection algorithm. As a complement to anomaly detection algorithms, it can achieve better results on problems that are no longer anomaly detection.