TY - UNPB
T1 - Reconstruction of Short-Lived Particles using Graph-Hypergraph Representation Learning
AU - Birch-Sykes, Callum
AU - Le, Brian
AU - Peters, Yvonne
AU - Simpson, Ethan
AU - Zhang, Zihan
N1 - 12 pages, 8 figures. To be submitted to Physical Review D (PRD)
PY - 2024/2/15
Y1 - 2024/2/15
N2 - In collider experiments, the kinematic reconstruction of heavy, short-lived particles is vital for precision tests of the Standard Model and in searches for physics beyond it. Performing kinematic reconstruction in collider events with many final-state jets, such as the all-hadronic decay of topantitop quark pairs, is challenging. We present HyPER, a graph neural network that uses blended graph-hypergraph representation learning to reconstruct parent particles from sets of final-state objects. HyPER is tested on simulation and shown to perform favorably when compared to existing state-of-the-art reconstruction techniques, while demonstrating superior parameter efficiency. The novel hypergraph approach allows the method to be applied to particle reconstruction in a multitude of different physics processes.
AB - In collider experiments, the kinematic reconstruction of heavy, short-lived particles is vital for precision tests of the Standard Model and in searches for physics beyond it. Performing kinematic reconstruction in collider events with many final-state jets, such as the all-hadronic decay of topantitop quark pairs, is challenging. We present HyPER, a graph neural network that uses blended graph-hypergraph representation learning to reconstruct parent particles from sets of final-state objects. HyPER is tested on simulation and shown to perform favorably when compared to existing state-of-the-art reconstruction techniques, while demonstrating superior parameter efficiency. The novel hypergraph approach allows the method to be applied to particle reconstruction in a multitude of different physics processes.
KW - hep-ph
M3 - Preprint
BT - Reconstruction of Short-Lived Particles using Graph-Hypergraph Representation Learning
ER -