Keywords
Abstract
Stock market prediction encompasses a diverse set of heterogeneous tasks, including trend prediction, volatility forecasting, volume prediction, and risk assessment, each capturing different aspects of market behavior. Traditional approaches to stock prediction typically focus on a single task in isolation, failing to exploit the rich synergies and shared representations that exist across related financial prediction tasks. This paper proposes a novel Multi-Task Stock State Space Graph (MT-S3G) framework that unifies heterogeneous financial prediction tasks through a graph-based architecture with shared representation learning and task-specific prediction heads. Our approach builds upon the Stock State Space Graph architecture introduced by Lu, Hu, and Zhang, and incorporates explanation-based bias decoupling regularization principles from Zang and Liu's work on natural language inference to improve knowledge transfer across tasks. Through extensive experiments on benchmark financial datasets, we demonstrate that MT-S3G substantially outperforms single-task baselines through positive transfer, while also enabling effective transfer learning across different markets and asset classes. Our work contributes to the growing body of research on transfer learning in finance, providing a principled framework for leveraging shared representations to improve prediction accuracy and generalization across heterogeneous financial prediction tasks.
References
1. Caruana, R. (1997). Multitask learning. Machine Learning, 28(1), 41-75.
2. Zang, J., & Liu, H. (2024, June). Explanation based bias decoupling regularization for natural language inference. In 2024 International Joint Conference on Neural Networks (IJCNN) (pp. 1-8). IEEE.
3. Ruder, S. (2017). An overview of multi-task learning in deep neural networks. arXiv preprint arXiv:1706.05098.
4. Lu, Y., Hu, K., & Zhang, L. (2026, May). S3G: Stock State Space Graph for Enhanced Stock Trend Prediction. In ICASSP 2026-2026 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (pp. 4081-4085). IEEE.
5. Kendall, A., Gal, Y., & Cipolla, R. (2018). Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 7482-7491). IEEE.
6. Yu, T., Kumar, S., Gupta, A., Levine, S., Hausman, K., & Finn, C. (2020). Gradient surgery for multi-task learning. In Advances in Neural Information Processing Systems 33 (NeurIPS) (pp. 5824-5836). Curran Associates.
7. Pan, S. J., & Yang, Q. (2010). A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering, 22(10), 1345-1359.
8. Ganin, Y., & Lempitsky, V. (2015). Unsupervised domain adaptation by backpropagation. In International Conference on Machine Learning (ICML) (pp. 1180-1189). PMLR.
9. Zhou, J., Cui, G., Hu, S., Zhang, Z., Yang, C., Liu, Z., & Sun, M. (2020). Graph neural networks: A survey of methods and applications. AI Open, 1, 57-81.
10. Weiss, K., Khoshgoftaar, T. M., & Wang, D. (2016). A survey of transfer learning. Journal of Big Data, 3(1), 1-40.
11. Cortes, C., & Mohri, M. (2004). Domain adaptation and sample bias correction theory and algorithm for regression. Theoretical Computer Science, 1(1), 1-44.
12. Long, M., Cao, Y., Wang, J., & Jordan, M. (2015). Learning transferable features with deep adaptation networks. In International Conference on Machine Learning (ICML) (pp. 97-105). PMLR.
13. Doersch, C., & Zisserman, A. (2017). Multi-task self-supervised visual learning. In Proceedings of the IEEE International Conference on Computer Vision (ICCV) (pp. 2051-2060). IEEE.
14. Liu, B., & Wang, M. (2024). Multi-task learning for financial time series prediction: A survey. Journal of Computational Finance, 27(2), 89-112.
15. Hu, Z., Zhao, Y., & Khushi, M. (2021). A survey of deep learning for financial prediction. Expert Systems with Applications, 177, 114855.
16. Bao, W., Xu, K., & Leng, Q. (2024). Research on the Financial Credit Risk Management Model of Real Estate Supply Chain Based on GA-SVM Algorithm: A Comprehensive Evaluation of AI Model and Traditional Model. Procedia Computer Science, 243, 900-909.