Speaker
Description
Identifying jets originating from the decay of highly boosted heavy particles in colliders plays a crucial
role in uncovering potential signs of physics beyond the Standard Model. Despite significant progress in
jet-origin classification algorithms—particularly graph neural networks—the rapidly increasing volume
of collider data and the demand for faster and more efficient processing present a major challenge for
their future applicability. In this study, quantum machine learning is investigated as a powerful tool to
address this challenge. For the first time, a quantum graph neural network is designed to identify
boosted jets arising from hadronic decays of the Z boson, receiving data without any physics-driven
preprocessing and relying solely on dimensionality reduction. The reduction is performed using a
convolutional autoencoder whose performance improves in the presence of added noise. The overall
efficiency of the framework is evaluated in two training schemes—separate and joint training of the
autoencoder and the quantum graph network—and the results are compared with the algorithm used in
the CMS experiment at CERN. The findings indicate the strong potential of the quantum algorithm to
reproduce the performance of classical methods.
Keywords: jet identification algorithms, graph neural networks, quantum machine learning