Deep Learning Based Impact Parameter Determination for the CBM Experiment

In this talk we presented a novel technique, based on Deep Learning, to determine the impact parameter of nuclear collisions at CBM experiment. PointNet Learning models are trained UrQMD followed by CBMRoot simulations Au+Au 10 AGeV reconstruct from raw experimental data such as hits particles in detector planes, tracks reconstructed or their combinations. The can perform fast, accurate, event-by-event determination heavy ion collision experiments. They shown outperform simple model which maps track multiplicity parameter. While conventional methods for centrality classification merely provide an expected distribution given class, predict 2–14 fm basis with mean error ?0.33 0.22 fm.