Boosting mono-jet searches with model-agnostic machine learning

TL;DR

This paper demonstrates that weakly supervised machine learning, specifically the CWoLa method, can significantly enhance the sensitivity of LHC mono-jet searches for new physics by utilizing low-level detector data without relying on specific models.

Contribution

It introduces the application of the CWoLa method to mono-jet searches, enabling model-agnostic analysis and improved discovery potential using simulated data.

Findings

Boosted the discovery potential of mono-jet searches

Demonstrated effectiveness with simulated dark matter data

Showed model-agnostic approach enhances sensitivity

Abstract

We show how weakly supervised machine learning can improve the sensitivity of LHC mono-jet searches to new physics models with anomalous jet dynamics. The Classification Without Labels (CWoLa) method is used to extract all the information available from low-level detector information without any reference to specific new physics models. For the example of a strongly interacting dark matter model, we employ simulated data to show that the discovery potential of an existing generic search can be boosted considerably.