b-tagging using Neural Network

The existence of heavy particles, such as Higgs bosons and top quarks, which have short lifetime, cannot be detected directly and is inferred by the existence of their decay products. The bottom quark, observed as a jet of particles (b-jet) in the detector, is a common decay product of heavy particles. Therefore, the identification of b-jets (b-tagging) in particle detectors is essential for studying the physical processes of these heavy particles. b-tagging algorithms reconstruct particle trajectories, formulate parameters from the observed data, and classify particle flavors based on these parameters. This paper demonstrates that better b-tagging performance can be achieved by using neural network models, as opposed to using b-tagging parameters alone, as classifiers.