핵심 개념
Introducing capsule layers in GNN architectures enhances quark-gluon tagging performance.
초록
Introduction
Advanced machine learning boosts object tagging.
Novel architectural modifications with capsule layers in GNNs.
Dataset
Quark-gluon dataset generation and high-level features description.
Model
LorentzNet block details and capsule block overview.
Classifier Performance
Comparison between LorentzNet and CapsLorentzNet for quark-gluon tagging.
Summary and Outlook
Proposed architectural modification's effectiveness demonstrated.
통계
"Our new architecture can enhance the performance of LorentzNet by 20 % for the quark-gluon tagging task."
"The LorentzNet model is trained on both datasets, comparing its performance to test the similarity between the generated and preexisting datasets."