Trees vs neural networks for enhancing tau lepton real-time selection in proton-proton collisions

Abstract

Abstract This paper introduces supervised learning techniques for real-time selection (triggering) of hadronically decaying tau leptons in proton-proton colliders. By implementing traditional machine learning decision trees and advanced deep learning models, such as Multi-Layer Perceptron or residual neural networks, visible improvements in performance compared to standard rule-based tau triggers are observed. We show how such an implementation may lower selection energy thresholds, thus increasing the sensitivity of searches for new phenomena in proton-proton collisions classified by low-energy tau leptons. Moreover, we analyze when it is better to use neural networks vs decision trees for tau triggers with conclusions relevant to other problems in physics.