Partition Pooling for Convolutional Graph Network Applications in Particle Physics

TL;DR

This paper introduces a partition pooling scheme for convolutional graph networks in particle physics, enabling more efficient and deeper models for event reconstruction with improved accuracy and reduced overfitting.

Contribution

The paper proposes a novel partition pooling method that adapts image pooling techniques to graph neural networks in particle physics, enhancing performance and resource efficiency.

Findings

Partition pooling improves reconstruction accuracy.

Pooling reduces overfitting in deep graph networks.

Resource efficiency allows for deeper, more effective models.

Abstract

Convolutional graph networks are used in particle physics for effective event reconstructions and classifications. However, their performances can be limited by the considerable amount of sensors used in modern particle detectors if applied to sensor-level data. We present a pooling scheme that uses partitioning to create pooling kernels on graphs, similar to pooling on images. Partition pooling can be used to adopt successful image recognition architectures for graph neural network applications in particle physics. The reduced computational resources allow for deeper networks and more extensive hyperparameter optimizations. To show its applicability, we construct a convolutional graph network with partition pooling that reconstructs simulated interaction vertices for an idealized neutrino detector. The pooling network yields improved performance and is less susceptible to overfitting than a similar network without pooling. The lower resource requirements allow the construction of a deeper network with further improved performance.