GPUs for Higgs boson data analysis at the LHC using the Matrix Element Method

Schouten, Doug; DeAbreu, Adam; Stelzer, Bernd

doi:10.3204/DESY-PROC-2014-05/20

Contribution to a conference proceedings/Internal Report

PUBDB-2015-05337

GPUs for Higgs boson data analysis at the LHC using the Matrix Element Method

Schouten, D. ; DeAbreu, A. ; Stelzer, B.

2015
Deutsches Elektronen-Synchrotron, DESY Hamburg

GPU Computing in High-Energy Physics, GPUHEP2014, Pisa, Italy, 10 Sep 2014 - 12 Sep 2014 Hamburg : Deutsches Elektronen-Synchrotron, DESY 109-118 (2015) [10.3204/DESY-PROC-2014-05/20]

This record in other databases:

Please use a persistent id in citations: doi:10.3204/DESY-PROC-2014-05/20

Report No.: DESY-PROC-2014-05

Abstract: The matrix element method utilizes ab initio calculations of probability densities as power- ful discriminants to extract small signals from large backgrounds at hadron collider exper- iments. The computational complexity of this method for final states with many particles and degrees of freedom sets it at a disadvantage compared to supervised classification methods such as decision trees, k nearest-neighbour, or neural networks. We will present a concrete implementation of the matrix element technique in the context of Higgs boson analysis at the LHC employing graphics processing units. Due to the intrinsic paral- lelizability of multidimensional phase space integration, dramatic speedups can be readily achieved, which makes the matrix element technique viable for general usage at collider experiments.

Keyword(s): talk: Pisa 2014/09/10 ; Higgs particle ; CERN LHC Coll ; higher-dimensional ; neural network ; programming ; microprocessor: graphics ; phase space ; background ; density