Automated Feature Engineering for Multi-Period Claim Experience in Dynamic Actuarial Risk Modeling
Keywords:
Automated Feature Engineering, Actuarial Risk Modeling, Multi-Period Claims, Dynamic Modeling, Predictive AnalyticsAbstract
Actuarial risk modeling increasingly relies on longitudinal claim experience to capture evolving policyholder behavior and portfolio dynamics. Traditional feature engineering in this domain often depends on manual specification, which limits scalability and may overlook complex temporal interactions. This study investigates automated feature engineering techniques for multi-period claim experience within dynamic actuarial risk modeling. The objective is to evaluate whether algorithmic feature construction can improve predictive accuracy, stability, and interpretability compared with conventional manual feature sets. A synthetic but realistic claim dataset spanning multiple observation periods is used, incorporating frequency, severity, reporting delay, and payment patterns. Automated feature generation methods, including deep feature synthesis and genetic programming, are applied to create candidate predictors. These features are then integrated into several dynamic modeling frameworks, such as gradient boosting, random forests, and regularized regression. Model performance is assessed through cross-validated discrimination, calibration, and error metrics. Results indicate that automated feature engineering yields consistent gains in discriminatory power and reduces prediction error relative to baseline manual features. The most valuable engineered features capture temporal trends, volatility, and interaction effects that are not easily specified by hand. Stability analysis further shows that a subset of automated features remains robust across resampling and time-based validation. The findings suggest that automated feature engineering can enhance dynamic actuarial risk models by uncovering latent multi-period claim patterns while maintaining acceptable interpretability. Practical implications include more efficient model development, improved risk segmentation, and better support for pricing and reserving decisions. However, careful validation is required to avoid overfitting and to ensure that engineered features align with actuarial intuition. This research contributes a structured evaluation framework and highlights opportunities for integrating automated feature construction into routine actuarial practice.References
1. X. Chen et al., “Neural feature search: A neural architecture for automated feature engineering,” in 2019 IEEE International Conference on Data Mining (ICDM), 2019, pp. 71–80.
2. B. Mahohoho, C. Chimedza, F. Matarise, and S. Munyira, “The Advanced Actuarial Data Science Based AI-Driven Solutions for Automated Loss Reserving Under IFRS 17 in Non-Life Insurance,” Science, Engineering and Technology, vol. 5, no. 2, pp. 75–115, 2025.
3. J. Adamczyk and F. Malawski, “Comparison of manual and automated feature engineering for daily activity classification in mental disorder diagnosis,” Computing & Informatics, vol. 40, no. 4, p. 850, 2021.
4. S. Dao, M. Chen, and L. Wang, “Automated Feature Engineering for Credit Risk Assessment: A Domain-Aware AutoML Approach.”
5. E. Carter, J. Anderson, O. Martinez, B. Walker, and C. Paul, “Automated Feature Engineering Using AI for Predictive Modeling in Financial Services,” 2021.
6. U. Khurana, D. Turaga, H. Samulowitz, and S. Parthasrathy, “Cognito: Automated feature engineering for supervised learning,” in 2016 IEEE 16th International Conference on Data Mining Workshops (ICDMW), 2016, pp. 1304–1307.
7. C. Balona, “ActuaryGPT: Applications of large language models to insurance and actuarial work,” British Actuarial Journal, vol. 29, p. e15, 2024.
8. K. Kuo, B. Crompton, and F. Logan, Deep Learning and Actuarial Experience Analysis. Compact, 2019.
9. M. Luteijn, J. Tam, and F. Fan, “Bridging transparency and predictive power: integrating explainable ML into actuarial modelling,” British Actuarial Journal, vol. 31, p. e10, 2026.
10. J. Zhang, H. Gao, Y. Liang, et al., “Scale-Separated Conditioning for Style-Encoder-Free Diffusion Stylization,” arXiv preprint arXiv:2608.19719, 2026.
11. J. T. Baillargeon, L. Lamontagne, and E. Marceau, “Mining actuarial risk predictors in accident descriptions using recurrent neural networks,” Risks, vol. 9, no. 1, p. 7, 2020.
12. R. Richman, “AI in actuarial science,” SSRN 3218082, 2018.
13. S. Brindöpke, “Actuarial Data Science,” in The Digital Journey of Banking and Insurance, Volume I: Disruption and DNA, Cham: Springer International Publishing, 2021, pp. 119–136.
14. K. Wang, P. Wang, and C. Xu, “Toward efficient automated feature engineering,” in 2023 IEEE 39th International Conference on Data Engineering (ICDE), 2023, pp. 1625–1637.
15. A. Taha, B. Cosgrave, and S. Mckeever, “Using feature selection with machine learning for generation of insurance insights,” Applied Sciences, vol. 12, no. 6, p. 3209, 2022.
16. N. Hu, “A Data Science Approach for Evaluating Experience Variables in Insurance Risk Modeling,” in *2026 IEEE International Conference on Information Reuse and Integration for Data Science (IRI)*, 2026, pp. 282–287.
17. E. Zdravevski et al., “Improving activity recognition accuracy in ambient-assisted living systems by automated feature engineering,” IEEE Access, vol. 5, pp. 5262–5280, 2017.
18. G. Zhu, S. Jiang, X. Guo, C. Yuan, and Y. Huang, “Evolutionary automated feature engineering,” in Pacific Rim International Conference on Artificial Intelligence, Cham: Springer Nature Switzerland, 2022, pp. 574–586.
19. C. S. Pareek, “DriftGuard: A Temporal Drift Detection Framework for Validating Long-Duration Insurance and Actuarial Models,” in *2026 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS)*, 2026, pp. 290–295.
20. T. Roberts and S. J. Tonna, *Risk Modeling: Practical Applications of Artificial Intelligence, Machine Learning, and Deep Learning*. Hoboken, NJ: John Wiley & Sons, 2022.
21. H. T. Lam, J. M. Thiebaut, M. Sinn, B. Chen, T. Mai, and O. Alkan, “One button machine for automating feature engineering in relational databases,” arXiv preprint arXiv:1706.00327, 2017.
22. B. Mahohoho, “Automated Actuarial Data Analytics-Based Inflation Adjusted Frequency Severity Loss Reserving Model,” Open Journal of Statistics, vol. 14, no. 3, pp. 341–393, 2024.
23. D. Kuizinienė et al., “A comparative study of feature selection and feature extraction methods for financial distress identification,” PeerJ Computer Science, vol. 10, p. e1956, 2024.
24. R. P. Bonidia et al., “BioAutoML: automated feature engineering and metalearning to predict noncoding RNAs in bacteria,” Briefings in Bioinformatics, vol. 23, no. 4, p. bbac218, 2022.
25. X. Li, Z. Guo, and R. Liu, “Task-Adaptive Calibration for Multimodal Long-Context Extension,” Research Square, preprint, Aug. 23, 2026. doi: 10.21203/rs.3.rs-10769418/v1.
26. R. Salazar, F. Neutatz, and Z. Abedjan, “Automated feature engineering for algorithmic fairness,” Proceedings of the VLDB Endowment, vol. 14, no. 9, pp. 1694–1702, 2021.
27. B. Ghojogh et al., “Feature selection and feature extraction in pattern analysis: A literature review,” arXiv preprint arXiv:1905.02845, 2019.
28. L. Li et al., “Learning a Data-Driven Policy Network for Pre-Training Automated Feature Engineering,” in International Conference on Learning Representations (ICLR), 2023.
29. Y. Ning, Z. Guo, X. Li, and W. Guo, “Reasoning-Aware Error-Bounded KV-Cache Compression and Sparse Attention for Long-Context LLMs,” Research Square, preprint, Sep. 2026. doi: 10.21203/rs.3.rs-10952127/v1.
30. Y. Lucas et al., “Towards automated feature engineering for credit card fraud detection using multi-perspective HMMs,” Future Generation Computer Systems, vol. 102, pp. 393–402, 2020.
31. B. Mahohoho, C. Chimedza, F. Matarise, and S. Munyira, “Artificial Intelligence-based Automated Actuarial pricing and underwriting model for the General Insurance Sector,” Open Journal of Statistics, vol. 14, no. 3, pp. 294–340, 2024.
32. K. Kuo, “Deeptriangle: A deep learning approach to loss reserving,” Risks, vol. 7, no. 3, p. 97, 2019.
33. S. Dey, D. Basak, and K. C. Mondal, “Advancement in Insurance Risk Prediction: A Review of Data-Driven Approaches,” Journal of Convergence in Technology and Management: Global Nexus, vol. 1, no. 2, pp. 71–76, 2025.
34. N. Pretorius and S. Marais, “The convergence of machine learning operations and actuarial control,” South African Actuarial Journal, vol. 25, no. 1, pp. C113–C135, 2025.
35. J. Wu, "A Methodological Framework for Training Tool-Aware AI Entertainers for Live Streaming," Artificial Intelligence and Digital Technology, vol. 3, no. 3, pp. 1-12, 2026.
