Role of Neural Networks in the Search of the Higgs Boson at LHC

TL;DR

This paper explores the application of neural network classifiers, both traditional and neural chip-based, to improve the discrimination of Higgs boson signals from background noise at the LHC, particularly around 200 GeV.

Contribution

It demonstrates the effectiveness of neural networks, including a neural chip, for offline and online Higgs signal analysis at the LHC, introducing new hardware for real-time processing.

Findings

Neural networks improve Higgs signal discrimination.

Neural chip Totem enables potential real-time analysis.

Effective classification at Higgs mass of 200 GeV.

Abstract

We show that neural network classifiers can be helpful to discriminate Higgs production from background at LHC in the Higgs mass range M= 200 GeV. We employ a common feed-forward neural network trained by the backpropagation algorithm for off-line analysis and the neural chip Totem, trained by the Reactive Tabu Search algorithm, which could be used for on-line analysis.