Learning to Remove Pileup at the LHC with Jet Images

Journal of Physics: Conference Series

Patrick T. Komiske^1¹  Benjamin Nachman^2²  Eric M. Metodiev^1³ 

[1]Center for Theoretical Physics, Massachusetts Institute of Technology, Cambridge, MA 02139, USA^1

[2]Department of Physics, Harvard University, Cambridge, MA 02138, USA^3

[3]Physics Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA^2

DOI  :  10.1088/1742-6596/1085/4/042010

学科分类：物理（综合）

来源: Institute of Physics Publishing Ltd.

【 摘 要 】

We present the Pileup Mitgation with Machine Learning (PUMML) algorithm for pileup removal at the Large Hadron Collider (LHC) based on the jet images framework using state-of-the-art machine learning techniques. We demonstrate that our algorithm outperforms existing methods on a wide range of jet observables up to pileup levels of 140 collisions per bunch crossing. We also investigate what aspects of the event our algorithms are utilizing by understanding the learned parameters of a simplified version of the model.

【 授权许可】

Unknown