Dynamic GNNs for Predicting Train Cancellations on the Dutch Railway Network: A Multi-Season Study of Environmental and Operational Factors

Digital Technologies Research and Applications

Article

Dynamic GNNs for Predicting Train Cancellations on the Dutch Railway Network: A Multi-Season Study of Environmental and Operational Factors

Downloads

https://doi.org/10.54963/dtra.v5i1.1709
Brakenhoff, B., Alsahag, A. M. M., & Ziabari, S. S. M. (2026). Dynamic GNNs for Predicting Train Cancellations on the Dutch Railway Network: A Multi-Season Study of Environmental and Operational Factors. Digital Technologies Research and Applications, 5(1), 32–52. https://doi.org/10.54963/dtra.v5i1.1709
Volume 5 Issue 1: March 2026 (in progress)
Copyright (c) 2026 Brent Brakenhoff, Ali Mohammed Mansoor Alsahag, Seyed Sahand Mohammadi Ziabari
Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright

The authors shall retain the copyright of their work but allow the Publisher to publish, copy, distribute, and convey the work.

License

Digital Technologies Research and Applications (DTRA)  publishes accepted manuscripts under Creative Commons Attribution 4.0 International (CC BY 4.0). Authors who submit their papers for publication by DTRA agree to have the CC BY 4.0 license applied to their work, and that anyone is allowed to reuse the article or part of it free of charge for any purpose, including commercial use. As long as the author and original source are properly cited, anyone may copy, redistribute, reuse, and transform the content.

Authors

  • Brent Brakenhoff

    Informatics Institute, University of Amsterdam, 1090 GH Amsterdam, The Netherlands
  • Ali Mohammed Mansoor Alsahag

    Informatics Institute, University of Amsterdam, 1090 GH Amsterdam, The Netherlands
  • Seyed Sahand Mohammadi Ziabari

    Informatics Institute, University of Amsterdam, 1090 GH Amsterdam, The Netherlands

Received: 16 October 2025; Revised: 4 November 2025; Accepted: 31 December 2025; Published: 30 January 2026

Cancellations on the Dutch Railway network are a common and unpredictable occurrence; however, little research has focused on predicting these cancellations. Previous studies on the Dutch railway system have primarily concentrated on delay prediction. For this regression task, models such as XGBoost, Random Forest, Long Short-Term Memory (LSTM), and Gradient Boosting Decision Tree have been shown to perform well. Graph neural network-based models have been used for regression tasks on other transportation networks. We propose a Dynamic Graph Neural Network (DGNN) combined with an LSTM network for binary classification of cancelled trajectories. We compare the model with baseline models on a seasonal split to compare the feature importance across different seasons. Model performance is gauged using paired t-tests on bootstrapped F1 scores. Additionally, Precision, Recall, Balanced Accuracy, and AUC are considered metrics for further comparison. The newly proposed features achieve mostly positive feature importance scores across the models. Amongst the evaluated models, the proposed DGNN and XGBoost outperform the baseline models. Overall, the models underperform with F1 scores no higher than 0.4. This paper provides insight into the influence of various weather and operational features on cancellations on the Dutch railway network, with the operational features proving insightful.

Keywords:

Railway Network DGNN Binary Classification Time Series Graph Neural Network LSTM XGBoost

References

  1. Rijden de Treinen. Available online: https://www.rijdendetreinen.nl/ (accessed on 18 November 2024).
  2. Rijden de Treinen. Statistics for 2024. Available online: https://www.rijdendetreinen.nl/en/statistics/2024 (accessed on 13 January 2025).
  3. Rijden de Treinen. Statistics for 2023. Available online: https://www.rijdendetreinen.nl/en/statistics/2023 (accessed on 13 January 2025).
  4. Kämper, M. Predicting Delayed Trajectories Using Network Features: A Study on the Dutch Railway Network. arXiv preprint 2025, arXiv:2507.11776. DOI: https://doi.org/10.48550/arXiv.2507.11776
  5. Lei, W.; Alves, L.G.A.; Amaral, L.A.N. Forecasting the Evolution of Fast-Changing Transportation Networks Using Machine Learning. Nat. Commun. 2022, 13, 4252. DOI: https://doi.org/10.1038/s41467-022-31911-2
  6. Dekker, M.M.; Panja, D.; Dijkstra, H.A.; et al. Predicting Transitions across Macroscopic States for Railway Systems. PLoS One 2019, 14, 1–26. DOI: https://doi.org/10.1371/journal.pone.0217710
  7. Kecman, P.; Goverde, R.M.P. Online Data-Driven Adaptive Prediction of Train Event Times. IEEE Trans. Intell. Transp. Syst. 2015, 16, 465–474. DOI: https://doi.org/10.1109/TITS.2014.2347136
  8. Li, Z.; Wen, C.; Hu, R.; et al. Near-Term Train Delay Prediction in the Dutch Railways Network. Int. J. Rail Transp. 2020, 9, 520–539. DOI: https://doi.org/10.1080/23248378.2020.1843194
  9. Zhang, Y.D.; Liao, L.; Yu, Q.; et al. Using the Gradient Boosting Decision Tree Algorithm for a Train Delay Prediction Model Considering the Delay Propagation Feature. Adv. Prod. Eng. Manag. 2021, 16, 285–296. DOI: https://doi.org/10.14743/apem2021.3.400
  10. Wen, C.; Mou, W.; Huang, P.; et al. A Predictive Model of Train Delays on a Railway Line. J. Forecast. 2020, 39, 470–488. DOI: https://doi.org/10.1002/for.2639
  11. Zheng, Y.; Yi, L.; Wei, Z. A Survey of Dynamic Graph Neural Networks. arXiv preprint 2024, arXiv:2404.18211. DOI: https://doi.org/10.48550/arXiv.2404.18211
  12. Luo, X.; Zhu, C.; Zhang, D.; et al. STG4Traffic: A Survey and Benchmark of Spatial-Temporal Graph Neural Networks for Traffic Prediction. arXiv preprint 2023, arXiv:2307.00495. DOI: https://doi.org/10.48550/arXiv.2307.00495
  13. Xue, J.; Tan, R.; Ma, J.; et al. Data Science in Transportation Networks with Graph Neural Networks: A Review and Outlook. Data Sci. Transp. 2025, 7, 10. DOI: https://doi.org/10.1007/s42421-025-00124-6
  14. Fu, Q.; Ding, S.; Zhang, T.; et al. Short-Term Train Arrival Delay Prediction: A Data-Driven Approach. Railw. Sci. 2024, 3, 514–529.
  15. Raghavendran, P.; Gochhait, S.; Gunasekar, T. Optimizing Load Dispatch Using Artificial Neural Networks and Fractional Weighted Models. J. Vis. Exp. 2025, 68811. DOI: https://doi.org/10.3791/68811
  16. Gori, M.; Monfardini, G.; Scarselli, F. A New Model for Learning in Graph Domains. In Proceedings of the 2005 IEEE International Joint Conference on Neural Networks, Montreal, QC, Canada, 31 July–4 August 2005. DOI: https://doi.org/10.1109/IJCNN.2005.1555942
  17. Scarselli, F.; Gori, M.; Tsoi, A.C.; et al. The Graph Neural Network Model. IEEE Trans. Neural Netw. 2009, 20, 61–80.
  18. Feng, Z.; Wang, R.; Wang, T.; et al. A Comprehensive Survey of Dynamic Graph Neural Networks: Models, Frameworks, Benchmarks, Experiments and Challenges. arXiv preprint 2024, arXiv:2405.00476. DOI: https://doi.org/10.48550/arXiv.2405.00476
  19. Veličković, P.; Cucurull, G.; Casanova, A.; et al. Graph Attention Networks. arXiv preprint 2018, arXiv:1710.10903. DOI: https://doi.org/10.48550/arXiv.1710.10903
  20. Furno, A.; Fanouzi, N.-E.E.; Sharma, R.; et al. Graph-Based Ahead Monitoring of Vulnerabilities in Large Dynamic Transportation Networks. PLoS One 2021, 16, 1–35. DOI: https://doi.org/10.1371/journal.pone.0248764
  21. Peng, H.; Du, B.; Liu, M.; et al. Dynamic Graph Convolutional Network for Long-Term Traffic Flow Prediction with Reinforcement Learning. Inf. Sci. 2021, 578, 401–416. DOI: https://doi.org/10.1016/j.ins.2021.07.007
  22. Kroev, E.; Koseoglu, B.; Traverso, L.; et al. Shap-Select: Lightweight Feature Selection Using SHAP Values and Regression. arXiv preprint 2024, arXiv:2410.06815. DOI: https://doi.org/10.48550/arXiv.2410.06815
  23. Altmann, A.; Toloşi, L.; Sander, O.; et al. Permutation Importance: A Corrected Feature Importance Measure. Bioinformatics 2010, 26, 1340–1347. DOI: https://doi.org/10.1093/bioinformatics/btq134
  24. Oneto, L.; Fumeo, E.; Clerico, G.; et al. Train Delay Prediction Systems: A Big Data Analytics Perspective. Big Data Res. 2018, 11, 54–64. DOI: https://doi.org/10.1016/j.bdr.2017.05.002
  25. Hochreiter, S.; Schmidhuber, J. Long Short-Term Memory. Neural Comput. 1997, 9, 1735–1780. DOI: https://doi.org/10.1162/neco.1997.9.8.1735
  26. Koninklijk Nederlands Meteorologisch Instituut. About KNMI. Available online: https://www.knmi.nl/nederland-nu/klimatologie/daggegevens (accessed on 15 January 2025).
  27. Chawla, N.V.; Bowyer, K.W.; Hall, L.O.; et al. SMOTE: Synthetic Minority Over-Sampling Technique. J. Artif. Intell. Res. 2002, 16, 321–357.
  28. Wu, J.; Chen, X.-Y.; Zhang, H.; et al. Hyperparameter Optimization for Machine Learning Models Based on Bayesian Optimization. J. Electron. Sci. Technol. 2019, 17, 26–40.