Study of novel pileup mitigation techniques in quark/gluon and tau jets using machine learning

The goal of this project is testing new Pileup Per Particle Identification (PUPPI) algorithm for identification and reconstruction of hadronic decays of tau leptons, alongside so far used algorithms with the same purpose, hadron-plus-strips (HPS) and Charged Hadron Subtraction (CHS) algorithm. Tests are done on the set of data for the next run, simulated by Monte Carlo (MC). This report presents preliminary results of these tests, showing efficiency and purity of algorithms for for several requirements on the tau identification.