Author(s)

Weiyi Zhu ¹, Yu Liu ¹, Siyuan Jiang ¹, Siyuan Pan ¹, Wen Zhong ¹

Affiliation(s)

¹ School of Big Data and Statistics, Sichuan Tourism University, Chengdu, Sichuan, China

Corresponding Author

Wen Zhong

ABSTRACT

Risk stratification on high-dimensional clinical tabular data poses fundamental challenges arising from feature heterogeneity, class imbalance, and the tension between predictive accuracy and interpretability required for principled decision support. This paper presents a two-stage ensemble learning framework that integrates gradient boosting machines with logistic regression baselines under a hybrid Bayesian-Grid hyperparameter optimization scheme reinforced by Bootstrap resampling validation. The proposed architecture employs systematic data preprocessing through statistical imputation, robust outlier detection, and feature standardization, followed by a three-model ensemble combining XGBoost, LightGBM, and regularized logistic regression. A hierarchical hyperparameter optimization pipeline fuses the global exploration capability of Bayesian optimization with the local refinement of grid search, while Bootstrap resampling with 500 iterations ensures parameter stability and provides confidence intervals on the resulting performance estimates. Strict separation between training and held-out test partitions preserves the integrity of generalization assessment to previously unseen patient records. Experimental evaluation on a clinical cohort of 1,247 records collected over a seven-year horizon achieves an overall risk stratification accuracy of 92.3%, substantially exceeding the 68.5% accuracy of the conventional single-marker baseline. The framework attains an area under the receiver operating characteristic curve of 0.946, an F1 score of 91.8%, and a recall of 90.1%, validating its practical utility for interpretable clinical decision support systems.

KEYWORDS

XGBoost Gradient Boosting, LightGBM Ensemble Learning, Bayesian Hyperparameter Optimization, Bootstrap Resampling Validation, Clinical Risk Stratification, Tabular Data Classification

CITE THIS PAPER

Weiyi Zhu, Yu Liu, Siyuan Jiang, Siyuan Pan, Wen Zhong. Stratified Bootstrap Validation and Bayesian-Grid Tuning for Robust Gradient Boosting Ensembles on Clinical Tabular Data. Information Systems and Economics (2026). Vol. 7, No.1, 51-61. DOI: http://dx.doi.org/10.23977/infse.2026.070106.

REFERENCES

[1] Alshayeji, M.H. and Abed, S.E. (2025) Heart disease prediction by tabular modeling with deep learning network and interpretability. Machine Learning: Science and Technology, 6, 035043.
[2] Wu, B., Ding, Z. and Huang, J. (2026) A review of continual learning in edge AI. IEEE Transactions on Network Science and Engineering.
[3] Jayakarthik, R., Gopinath, D., Begum, M.A., Fuladi, A.D., Balaram, A. and Raja, C. (2025) EnvHealthNet: A multi-modal machine learning model for commercial environmental health risk prediction. Proceedings of the 2025 5th International Conference on Pervasive Computing and Social Networking (ICPCSN), 1121-1128.
[4] Wu, B., Ding, Z., Ostigaard, L. and Huang, J. (2025) Reinforcement learning-based energy-aware coverage path planning for precision agriculture. Proceedings of the 2025 ACM Research on Adaptive and Convergent Systems (RACS), 1-8.
[5] Abasi, A., Nazari, A., Moezy, A. and Fatemi Aghda, S.A. (2025) Machine learning models for reinjury risk prediction using cardiopulmonary exercise testing (CPET) data: Optimizing athlete recovery. BioData Mining, 18, 16.
[6] Wu, B., Cai, Z., Wu, W. and Yin, X. (2023) AoI-aware resource management for smart health via deep reinforcement learning. IEEE Access, 11, 81180-81195.
[7] Leckey, C., Van Dyk, N., Doherty, C., Lawlor, A. and Delahunt, E. (2025) Machine learning approaches to injury risk prediction in sport: A scoping review with evidence synthesis. British Journal of Sports Medicine, 59, 491-500.
[8] Wu, B. and Wu, W. (2023) Model-free cooperative optimal output regulation for linear discrete-time multi-agent systems using reinforcement learning. Mathematical Problems in Engineering, 6350647.
[9] Lazar, A., Pokhrel, P. and Das, S. (2026) Beyond accuracy: A comprehensive comparative study of gradient boosting versus tabular deep learning and explainability techniques for mixed-type tabular data models using SHAP and LIME. International Journal on Artificial Intelligence Tools, 35, 2640003.
[10] Yıldız, A.Y. and Kalayci, A. (2025) Gradient boosting decision trees on medical diagnosis over tabular data. Proceedings of the 2025 IEEE International Conference on AI and Data Analytics (ICAD), 1-8.
[11] Karagoz, G., Ozcelebi, T. and Meratnia, N. (2025) Systematic benchmarking of local and global explainable AI methods for tabular healthcare data. Proceedings of the World Conference on Explainable Artificial Intelligence, Springer Nature Switzerland, 337-358.
[12] Champahom, T., Banyong, C., Janhuaton, T., Se, C., Watcharamaisakul, F., Ratanavaraha, V. and Jomnonkwao, S. (2025) Deep learning vs. gradient boosting: Optimizing transport energy forecasts in Thailand through LSTM and XGBoost. Energies, 18, 1685.
[13] Ferreira, P., Martins, E., Silva, J. and Teixeira, P. (2025) Feature selection and XGBoost for enhanced intrusion detection: A comparative study across benchmark datasets. Proceedings of the 2025 13th International Symposium on Digital Forensics and Security (ISDFS), 1-6.
[14] Huang, J., Wu, B., Duan, Q., Dong, L. and Yu, S. (2025) A fast UAV trajectory planning framework in RIS-assisted communication systems with accelerated learning via multithreading and federating. IEEE Transactions on Mobile Computing.
[15] Kumar, R., Singhal, N. and Chhabra, A. (2025) Hybrid optimization algorithm with the combination of PSO and genetic algorithm for task scheduling in cloud computing. E-Learning and Digital Media, 20427530251331082.
[16] Nathiya, N., Rajan, C. and Geetha, K. (2025) A hybrid optimization and machine learning based energy-efficient clustering algorithm with self-diagnosis data fault detection and prediction for WSN-IoT application. Peer-to-Peer Networking and Applications, 18, 13.
[17] Wu, B., Huang, J. and Yu, S. (2026) 'X of Information' continuum: A survey on AI-driven multi-dimensional metrics for next-generation networked systems. IEEE Communications Surveys & Tutorials.
[18] Wu, B., Huang, J., Duan, Q., Dong, L. and Cai, Z. (2025) Enhancing vehicular platooning with wireless federated learning: A resource-aware control framework. IEEE/ACM Transactions on Networking, 33, 1-16.
[19] Hajihosseinlou, M., Maghsoudi, A. and Ghezelbash, R. (2025) A semi-supervised approach for mineral prospectivity mapping via weighted positive-unlabeled learning and tree-structured Parzen estimator for hyperparameter optimization. Ore Geology Reviews, 106783.
[20] Wu, B., Huang, J. and Duan, Q. (2025) FedTD3: An accelerated learning approach for UAV trajectory planning. Proceedings of the International Conference on Wireless Artificial Intelligent Computing Systems and Applications (WASA), 13-24.
[21] Zlobin, M. and Bazylevych, V. (2025) Bayesian optimization for tuning hyperparameters of machine learning models: A performance analysis in XGBoost. Computer Systems and Information Technologies, 1, 141-146.
[22] Khan, H., Khan, A., Villar, S., Alonso, L., Almaleh, A. and Al-Qahtani, A. (2025) A comparative study of optimized-LSTM models using tree-structured Parzen estimator for traffic flow forecasting in intelligent transportation. Computers, Materials & Continua, 83, 3369.
[23] Wu, B., Huang, J. and Duan, Q. (2025) Real-time intelligent healthcare enabled by federated digital twins with AoI optimization. IEEE Network, 1.
[24] Foggetti, A., Nucci, F. and Papadia, G. (2025) Tuning metaheuristics with tree-structured Parzen estimator: A case study on scheduling. Journal of Artificial Intelligence and Autonomous Intelligence, 2, 293-321.
[25] Pan, D., Wu, B.-N., Sun, Y.-L. and Xu, Y.-P. (2023) A fault-tolerant and energy-efficient design of a network switch based on a quantum-based nano-communication technique. Sustainable Computing: Informatics and Systems, 37, 100827.
[26] Agrawal, S.K. (2026) Adaptive density-aware clustering of high-dimensional patient data in electronic health records. International Journal of Engineering Development and Research, 14, 361-367.

Sponsors, Associates, and Links

---

• Accounting, Auditing and Finance
  
  
• Industrial Engineering and Innovation Management
  
  
• Tourism Management and Technology Economy
  
  
• Journal of Computational and Financial Econometrics
  
  
• Financial Engineering and Risk Management
  
  
• Accounting and Corporate Management
  
  
• Social Security and Administration Management
  
  
• Population, Resources & Environmental Economics
  
  
• Statistics & Quantitative Economics
  
  
• Agricultural & Forestry Economics and Management
  
  
• Social Medicine and Health Management
  
  
• Land Resource Management
  
  
• Information, Library and Archival Science
  
  
• Journal of Human Resource Development
  
  
• Manufacturing and Service Operations Management
  
  
• Operational Research and Cybernetics