2512.04839

Total: 1

#1 Optimal Transport Event Representation for Anomaly Detection [PDF1] [Copy] [Kimi1] [REL]

Authors: Aditya Bhargava, Tianji Cai, Benjamin Nachman

We introduce optimal transport (OT) as a physics-based intermediate event representation for weakly supervised anomaly detection. With only $0.5\%$ injection of resonant signals in the LHC Olympics benchmark datasets, the OT-augmented feature set achieves nearly twice the significance improvement of standard high-level observables, while end-to-end deep learning on low-level four-momenta struggles in the low-signal regime. The gains persist across signal types and classifiers, underscoring the value of structured representations in machine learning for anomaly detection.

Subjects: High Energy Physics - Phenomenology , High Energy Physics - Experiment

Publish: 2025-12-04 14:21:50 UTC