Quantum generative adversarial networks for high energy physics
Date
2025-04
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Benemérita Universidad Autónoma de Puebla
Abstract
"The increasing computational demands of the High-Luminosity Large Hadron Collider (HL-LHC) have made quantum computing a promising tool for advancing high-energy physics (HEP) research. This thesis explores the application of Quantum Generative Adversarial Networks (QGANs) to address two central challenges in HEP: particle identification and data generation. Specifically, a dual-task QGAN model was developed using simulated data from Delphes, aiming both to distinguish real from generated events and to classify jets as either signal or background. The model integrates quantum generators and discriminators through TensorFlow Quantum and Google Cirq, thus enabling a hybrid architecture that combines quantum circuits with classical neural networks. As a result, the system achieved a classification accuracy of 90%, showing promising capabilities in mimicking real data distributions while simultaneously performing classification. However, it is important to note that this work does not claim quantum advantage, as the technology is still at an early stage. Nevertheless, the study demonstrates the feasibility of applying QGANs to computationally demanding HEP problems and suggests that, with further exploration into circuit design, data complexity, and scalability, quantum computing could serve as a complementary method to classical approaches in the HL-LHC era".
Description
Keywords
Citation
Collections
Document Viewer
Select a file to preview:
Can't see the file? Try reloading