Tandem Project Report: Classification in particle physics using machine learning
Published in Tandem Project Reports, 2020
Recommended citation: G. Bianco et. al, "Tandem Project Report: Classification in particle physics using machine learning" (2020).
The iTHEPHY Tandem Project is designed for students from Italy, Germany and France, where small groups of students are built and supervised by postdoctoral researchers and professors. In the context of particle physics, different experimental and theoretical topics are treated. Our work was to get an introduction to machine learning techniques and to understand a measurement by the ATLAS collaboration about the coupling of Higgs boson to tau leptons. With regard to that, we performed a machine learning classification on a subset from the ATLAS analysis, simulated using Monte Carlo methods at a centre-of-mass energy of 8 TeV. We tried two complementary machine learning approaches: Deep Neural Networks and Boosted Decision Trees. The combination of both outputs was found to give the most significant separation of signal and background events.