Quark-Gluon-Classification

Classification of Quarks and Gluons using Graph Neural Networks with Pooling (with PyTorch Geometric)

We aim to train a classifier of Quarks and Gluons (possibly Electrons/Photons later). For that we use Graph Neural Networks. More specifically, GraphSAGE message passing layers and EdgePooling operations to compress the graph representation and learn a classification score based on node features.