End-to-End Multi-Track Reconstruction using Graph Neural Networks at Belle II

TL;DR

This paper introduces a novel end-to-end graph neural network approach for multi-track reconstruction in the Belle II experiment, significantly improving efficiency and accuracy over traditional methods in realistic collision environments.

Contribution

It presents the first end-to-end multi-track ML algorithm for drift chambers in a realistic physics setting, enhancing track reconstruction performance.

Findings

Achieves 85.4% efficiency for complex event topologies.

Reduces fake rate to 2.5%.

Outperforms baseline algorithms in realistic simulations.

Abstract

We present the study of an end-to-end multi-track reconstruction algorithm for the central drift chamber of the Belle II experiment at the SuperKEKB collider using Graph Neural Networks for an unknown number of particles. The algorithm uses detector hits as inputs without pre-filtering to simultaneously predict the number of track candidates in an event and their kinematic properties. In a second step, we cluster detector hits for each track candidate to pass to a track fitting algorithm. Using a realistic full detector simulation including beam-induced backgrounds and detector noise taken from actual collision data, we find significant improvements in track finding efficiencies for tracks in a variety of different event topologies compared to the existing baseline algorithm used in Belle II. For events with a hypothetical long-lived massive particle with a mass in the GeV-range decaying uniformly along its flight direction into two charged particles, the GNN achieves a combined track finding and fitting efficiency of 85.4% with a fake rate of 2.5%, compared to 52.2% and 4.1% for the baseline algorithm. This is the first end-to-end multi-track machine learning algorithm for a drift chamber detector that has been utilized in a realistic particle physics environment.