One of the significant tasks (at the offline analysis stage) during HEP experiments is charged particle identification (PID). Over the last ten years, machine learning approaches have become widely used in high energy physics problems in general and in PID in particular. This work is devoted to the machine learning application for PID at the MPD and BM@N experiments. Numerical computations of the study are conducted in the ecosystem for tasks of machine learning, deep learning and data analysis.
