Cut-Based Photon ID Tuning of CMS Using Genetic Algorithm

Whenever we expect a photon as a final state particle of a collider search, it is essential to identify whether the photon is coming from a hard scattering at the primary vertex of the collision(prompt photons) or from decays of \(\pi ^0\) and fragmentation processes within a jet depositing their energies in the electromagnetic calorimeter. The goal of photon identification(ID) is to accept prompt photons at a high efficiency while rejecting the non-prompt photons maximally, for a given efficiency. Identifying prompt photons and measuring the energy deposited by the photons accurately, play a crucial role for analyses which have photon(s) in the final state. In this talk, we will discuss the optimization of the photon identification criteria obtained by using genetic algorithm. Using Monte Carlo samples of CMS corresponding to run conditions of 2017, different working points of the ID corresponding to signal efficiencies \(90\%\), \(80\%\), and \(70\%\) are derived for both barrel and endcap.