Machine Learning Application for Λ Hyperon Reconstruction in CBM at FAIR

Shahid Khan,Andrea Dubla,Ilya Selyuzhenkov,Ali Imdad Khan,Olha Lavoryk,Oleksii Lubynets,Viktor Klochkov,G David,T Ullrich,I.-K Yoo,S Mukherjee,A Kalweit,Z Xu,P Garg

doi:10.1051/epjconf/202225913008

Abstract

The Compressed Baryonic Matter experiment at FAIR will investigate the QCD phase diagram in the region of high net-baryon densities. Enhanced production of strange baryons, such as the most abundantly produced Λ hyperons, can signal transition to a new phase of the QCD matter. In this work, the CBM performance for reconstruction of the Λ hyperon via its decay to proton and π− is presented. Decay topology reconstruction is implemented in the Particle-Finder Simple (PFSimple) package with Machine Learning algorithms providing effcient selection of the decays and high signal to background ratio.

Highlights

Theoretical calculations predict the possibility of a first order phase transition from hadron gas to a deconfined phase of strongly interacting matter and the existence of a critical point in the region of the QCD phase diagram above 450 MeV chemical potential and below 180 MeV temperature [1]
The Compressed Baryonic Matter (CBM) performance for reconstruction of the Λ hyperon via its decay to proton and π− using Machine Learning algorithms to achieve an efficient selection of the decays and high signal to background ratio is presented
We study Machine Learning (ML) performance using the same variables as used by the Kalman Filter Particle Finder (KFPF)

Summary

Introduction

Theoretical calculations predict the possibility of a first order phase transition from hadron gas to a deconfined phase of strongly interacting matter and the existence of a critical point in the region of the QCD phase diagram above 450 MeV chemical potential (μB) and below 180 MeV temperature [1]. The CBM performance for reconstruction of the Λ hyperon via its decay to proton and π− using Machine Learning algorithms to achieve an efficient selection of the decays and high signal to background ratio is presented. Cellular automaton and Kalman Filter Particle (KFParticle) based PFSimple package [7] are used for track finding, fitting and decay kinematics reconstruction.

Results

Conclusion