Unsupervised clustering for collider physics

2020/10/31 by V. M. Mikuni, Vinicius Mikuni, F. Canelli +1
Computer Science · Physics and Astronomy · #Anomaly detection #Artificial intelligence #Artificial neural network #Cluster analysis #Collider #Computational Physics and Python Applications #Computer science #Embedding #Machine learning #Particle Detector Development and Performance #Particle physics #Particle physics theoretical and experimental studies #Physics #Space (punctuation) #Unsupervised learning #hep-ex #physics.data-an

paper · pdf · doi:10.1103/physrevd.103.092007

published as Phys. Rev. D 103, 092007 (2021)

openalex publication_date 2021/05/24 · arxiv created 2021/05/31 · arxiv updated 2021/06/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/06

Abstract

We propose a new method for unsupervised clustering for collider physics named UCluster, where information in the embedding space created by a neural network is used to categorize collision events into different clusters that share similar properties. We show how this method can be developed into an unsupervised multiclass classification of different processes and applied in the anomaly detection of events to search for new physics phenomena at colliders.

Citations

• Adam: A Method for Stochastic Optimization
• The anti-ktjet clustering algorithm
• Search for new particles decaying into dijets in proton-antiproton collisions ats=1.96 TeV
• Search for New Phenomena in Dijet Mass and Angular Distributions from pp Collisions at √(s) = 13 TeV with the ATLAS Detector
• The Hungarian method for the assignment problem
• Fast inference of deep neural networks in FPGAs for particle physics
• Review of Particle Physics
• PLIERS at TREC8 - Appendix