Performance of FARICH-based PID in realistic environment using ML

A detailed study of particle identification by the Focusing Aerogel Ring Imaging Cherenkov subsystem at the future charm superfactory detector is presented. To mitigate complex geometric distortions, a dedicated ring reconstruction algorithm was implemented in the detector simulation, considering realistic operating conditions. The algorithm’s performance was tested using single particles generated within the Aurora framework. To maximize separation power, two boosted decision tree-based classifiers for particle identification were developed for moderate and the most conservative assumptions about photosensor noise levels, scaling up to the dark count rate of [Formula: see text]. The approach is validated with the analysis of the [Formula: see text] decays, demonstrating that the systematic uncertainty and background contribution related to the [Formula: see text] separation performance can be minimized to strictly below [Formula: see text] owing to the high efficiency and robustness of the developed machine learning particle identification algorithm.