ML-based classification of photons and neutral mesons for Pb–Pb collisions recorded by ALICE

Direct photons are unique probes to study and characterize the quark-gluon plasma (QGP) as they are produced at all stages of heavy-ion collisions and leave the collision medium practically unscathed. Measurements at top Large Hadron Collider (LHC) energies at low p T reveal a very small thermal photon signal accompanied by considerable systematic uncertainties. Reduction of such uncertainties, which arise from the π 0 and η measurements and the photon identification, is crucial for obtaining significant results to be compared to the available theoretical calculations. To address these challenges, a novel approach employing machine learning (ML) techniques has been implemented for the classification of photons. An open-source set of frameworks comprising hipe4ml and scikit-learn packages is chosen for training, validation, and testing the model on a part of Run 2 Pb–Pb data at √ s NN = 5.02 TeV collision energy. In this paper, the performance of the novel approach in comparison to the standard cut-based analysis is presented. Initial findings employing gradientboosted decision trees demonstrate a substantial enhancement in photon purity while preserving efficiency levels comparable to those of the standard cut-based method. Strategies for addressing highly imbalanced data sets, including techniques like feature reduction during training and the implementation of scaled penalty factors to enhance discrimination between signal and background are also addressed. Finally, the feasibility of incorporating such ML methods into the main workflow of direct photon analysis is also presented