Identifying Jets Using Linear Discriminant Analysis

POSTER

Abstract

In order to shed light on the nature of Quark Gluon Plasma (QGP), we utilize the same machine learning principles that are used in image recognition to distinguish among different types of jets, namely jets from PYTHIA-generated proton-proton collisions and quenched jets from heavy ion collisions. We represent jet data as pixelated images, and these jet images are run through a series of preprocessing steps so as to standardize them as much as possible. Next, we use a Linear Discriminant Analysis (LDA) to distinguish between these jet images. The LDA is able to discern quite well among jets that come from different types of collisions.

Authors

  • Sofia Checa

    Yale Univ

  • Helen Caines

    Yale Univ

  • Benjamin Rosand

    Yale University, Yale Univ