![[2602.14571] DCTracks: An Open Dataset for Machine Learning-Based Drift Chamber Track Reconstruction](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.14571] DCTracks: An Open Dataset for Machine Learning-Based Drift Chamber Track Reconstruction
Summary
The article presents DCTracks, a new open dataset designed for machine learning-based track reconstruction in drift chambers, featuring standardized metrics for evaluation.
Why It Matters
This dataset is significant as it provides a foundation for advancing machine learning techniques in high energy physics, enabling researchers to compare traditional and modern algorithms effectively. It fosters reproducibility and innovation in track reconstruction methodologies.
Key Takeaways
- Introduction of a new Monte Carlo dataset for drift chamber events.
- Standardized metrics defined for evaluating track reconstruction methods.
- Facilitates comparison between traditional algorithms and Graph Neural Networks.
Computer Science > Machine Learning arXiv:2602.14571 (cs) [Submitted on 16 Feb 2026] Title:DCTracks: An Open Dataset for Machine Learning-Based Drift Chamber Track Reconstruction Authors:Qian Liyan, Zhang Yao, Yuan Ye, Zhang Zhaoke, Fang Jin, Jiang Shimiao, Zhang Jin, Li Ke, Liu Beijiang, Xu Chenglin, Zhang Yifan, Jia Xiaoqian, Qin Xiaoshuai, Huang Xingtao View a PDF of the paper titled DCTracks: An Open Dataset for Machine Learning-Based Drift Chamber Track Reconstruction, by Qian Liyan and 12 other authors View PDF HTML (experimental) Abstract:We introduce a Monte Carlo (MC) dataset of single- and two-track drift chamber events to advance Machine Learning (ML)-based track reconstruction. To enable standardized and comparable evaluation, we define track reconstruction specific metrics and report results for traditional track reconstruction algorithms and a Graph Neural Networks (GNNs) method, facilitating rigorous, reproducible validation for future research. Subjects: Machine Learning (cs.LG); High Energy Physics - Experiment (hep-ex) Cite as: arXiv:2602.14571 [cs.LG] (or arXiv:2602.14571v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2602.14571 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Liyan Qian [view email] [v1] Mon, 16 Feb 2026 09:04:49 UTC (2,967 KB) Full-text links: Access Paper: View a PDF of the paper titled DCTracks: An Open Dataset for Machine Learning-Based Drift Chamber Track Reconstructi...