Performance of the Vector Finder Toolkit for Track Reconstruction at the BM@N Experiment

The Baryonic Matter at Nuclotron (BM@N) experiment at the NICA accelerator complex is currently in its active phase, analyzing data collected from the first physics run, and preparing for future data taking. One of the most important software components for event reconstruction in high-multiplicity environment of heavy-ion interactions is a track reconstruction package. At BM@N, it was developed and implemented as a Vector Finder software toolkit, based on a constrained combinatorial search approach. The current status of the method performance is presented for simulated and real data from the BM@N experiment. To improve reconstruction quality, a fake track suppression method based on machine learning techniques using ROOT TMVA package, was implemented for BM@N and its characteristics are demonstrated.