A Study on Predicting Adolescent Cross-Sport Performance Using Shared Feature Space and Multi-Task Regression

Article Sidebar

PDF
Published: 2026-02-13
Keywords:
cross-sport prediction, multi-task learning, shared feature space, athletic performance, LightGBM

Main Article Content

Yue Hou
College of Engineering, Northeastern University, Boston, MA, 02115, United States

Abstract

Common transfer patterns exist across different sports disciplines in adolescents regarding abilities such as speed, explosiveness, and rhythm control. This study constructs a cross-discipline prediction framework based on shared feature space, targeting throwing velocity, sprint performance, and jumping ability. The framework comprises a general ability feature layer, discipline-specific feature layer, and multi-task regression head. General features include motion rhythm statistics, velocity distribution, acceleration trends, and training load indicators, while the discipline layer incorporates discipline-specific motion information. Predictions are performed using multi-head LightGBM. Experiments on 2.1 million sequence data points from 619 athletes demonstrate an average MAPE of 6.5%, representing a 22.7% improvement over single-task models. The framework maintains stable advantages even under sparse data conditions, indicating that the shared feature structure effectively enhances cross-sport prediction capabilities for adolescent athletic performance.

Article Details

Issue

Vol. 2 No. 1 (2026)

Section

Articles

How to Cite

A Study on Predicting Adolescent Cross-Sport Performance Using Shared Feature Space and Multi-Task Regression. (2026). Journal of Science, Innovation & Social Impact, 2(1), 124-131. https://sagespress.com/index.php/JSISI/article/view/86

References

1. X. Zheng, V. M. Dwyer, L. A. Barrett, M. Derakhshani, and S. Hu, "Rapid vital sign extraction for real-time opto-physiological monitoring at varying physical activity intensity levels," IEEE Journal of Biomedical and Health Informatics, vol. 27, no. 7, pp. 3107-3118, 2023. doi: 10.1109/jbhi.2023.3268240

2. M. Molnár, "Sport seasons through the lenses of project management: Identifying the common key characteristics of sport seasons and projects," Vezetéstudomány-Budapest Management Review, vol. 55, no. 7-8, pp. 32-44, 2024.

3. N. A. Bhat, and S. U. Farooq, "An empirical evaluation of defect prediction approaches in within-project and cross-project context," Software Quality Journal, vol. 31, no. 3, pp. 917-946, 2023. doi: 10.1007/s11219-023-09615-7

4. N. A. Bhat, and S. U. Farooq, "An improved method for training data selection for cross-project defect prediction," Arabian Journal for Science and Engineering, vol. 47, no. 2, pp. 1939-1954, 2022. doi: 10.1007/s13369-021-06088-3

5. S. Tang, S. Huang, and C. Zheng, "A novel cross-project software defect prediction algorithm based on transfer learning," Tsinghua Science and Technology, vol. 27, no. 1, pp. 41-57, 2021. doi: 10.26599/tst.2020.9010040

6. K. E. Bennin, A. Tahir, and S. G. MacDonell, "An empirical study on the effectiveness of data resampling approaches for cross-project software defect prediction," IET Software, vol. 16, no. 2, pp. 185-199, 2022.

7. S. Panigrahi, N. Stewart, and C. Sripada, "Selective inference for sparse multitask regression with applications in neuroimaging," The Annals of Applied Statistics, vol. 18, no. 1, pp. 445-467, 2024. doi: 10.1214/23-aoas1796

8. D. Kodati, and R. Tene, "Advancing mental health detection in texts via multi-task learning with soft-parameter sharing transformers," Neural Computing and Applications, vol. 37, no. 5, pp. 3077-3110, 2025. doi: 10.1007/s00521-024-10753-7

9. C. E. Tomlinson, P. J. Laurienti, and R. G. Lyday, "3M_BANTOR: A regression framework for multitask and multisession brain network distance metrics," Network Neuroscience, vol. 7, no. 1, pp. 1-21, 2023. doi: 10.1162/netn_a_00274

10. B. Patel, A. Orlichenko, and A. Patel, "Explainable multimodal graph isomorphism network for interpreting sex differences in adolescent neurodevelopment," Applied Sciences, vol. 14, no. 10, p. 4144, 2024. doi: 10.3390/app14104144