Beyond Cuts in Small Signal Scenarios -- Enhanced Sneutrino Detectability Using Machine Learning

TL;DR

This paper demonstrates that machine learning techniques, specifically XGBoost and deep neural networks, can significantly improve the detection sensitivity of supersymmetric sneutrino signals at the LHC compared to traditional methods.

Contribution

It introduces a machine learning-based approach with analysis techniques like template fitting and Shapley decomposition for enhanced new physics searches at the LHC.

Findings

Machine learning models outperform cut-and-count methods in signal detection.

Template fit yields better sensitivity than simple cut-based analysis.

Shapley decomposition provides insights into model decision-making.

Abstract

We investigate enhancing the sensitivity of new physics searches at the LHC by machine learning in the case of background dominance and a high degree of overlap between the observables for signal and background. We use two different models, XGBoost and a deep neural network, to exploit correlations between observables and compare this approach to the traditional cut-and-count method. We consider different methods to analyze the models' output, finding that a template fit generally performs better than a simple cut. By means of a Shapley decomposition, we gain additional insight into the relationship between event kinematics and the machine learning model output. We consider a supersymmetric scenario with a metastable sneutrino as a concrete example, but the methodology can be applied to a much wider class of models.