Latency-Aware Risk Control for Budget-Constrained Payment Fraud Decisioning

Bingjie Zi; Jin Wang; Yuchen Zhang

doi:10.6911/

Authors

Bingjie Zi Northeastern University, Boston, USA
Jin Wang University of California, Los Angeles, USA
Yuchen Zhang Northeastern University, Boston, USA

DOI:

https://doi.org/10.6911/

Keywords:

Payment fraud, risk decisioning, conformal risk control, verification latency, cost-sensitive learning, chargebacks.

Abstract

Production payment-risk engines do not merely score transactions; they must convert scores into approve/review actions under a fixed manual-review capacity, while the ground truth they are calibrated on arrives weeks later through chargebacks. We formulate fraud decisioning as monetary risk control and present LARC, a calibration layer that bounds the expected fraud value released to auto-approval at a user-specified level. LARC combines three components: a triage index ordered by expected fraud value rather than by score, which we show is optimal for monetary risk under a uniform review cost; amount-stratified conformal risk control whose per-stratum budgets are allocated by Lagrangian water-filling, made feasible on heavy-tailed amounts by an exposure cap that renders the monetary loss bounded; and an inverse-probability-of-censoring correction that removes the downward bias induced by immature chargeback labels. Experiments on 284,807 real card transactions and a 24-month simulated transaction stream show that every baseline we test—fixed alert rates, cost-sensitive Bayes thresholds, count-based conformal control and uncorrected conformal risk control—exceeds the risk target, whereas LARC holds it on both datasets and attains the highest net value among valid policies. Stratification cuts the review rate by up to 72% at equal risk, value ordering cuts realised risk by 25–29% at an equal review rate, and the latency correction keeps the guarantee intact down to 30% label maturity, where uncorrected calibration overshoots the target by 3.3×.

Downloads

Download data is not yet available.

References

[1] Angelopoulos, A. N., Bates, S., Fisch, A., Lei, L., & Schuster, T. (2024). Conformal risk control. In Proceedings of the International Conference on Learning Representations.

[2] Dal Pozzolo, A., Boracchi, G., Caelen, O., Alippi, C., & Bontempi, G. (2018). Credit card fraud detection: A realistic modeling and a novel learning strategy. IEEE Transactions on Neural Networks and Learning Systems, 29(8), 3784–3797. https://doi.org/10.1109/TNNLS.2017.2766186

[3] Carcillo, F., Dal Pozzolo, A., Le Borgne, Y.-A., Caelen, O., Mazzer, Y., & Bontempi, G. (2018). SCARFF: A scalable framework for streaming credit card fraud detection with Spark. Information Fusion, 41, 182–194. https://doi.org/10.1016/j.inffus.2017.09.001

[4] Correa Bahnsen, A., Aouada, D., Stojanovic, A., & Ottersten, B. (2016). Feature engineering strategies for credit card fraud detection. Expert Systems with Applications, 51, 134–142. https://doi.org/10.1016/j.eswa.2015.12.030

[5] Elkan, C. (2001). The foundations of cost-sensitive learning. In Proceedings of the 17th International Joint Conference on Artificial Intelligence (pp. 973–978).

[6] Chen, J., Liang, Y., Liu, J., & Zhou, M. (2026). Temporal transformer with conditional tabular GAN for credit card fraud detection: A sequential deep learning approach. Mathematics, 14(7), Article 1183. https://doi.org/10.3390/math14071183

[7] Vovk, V., Gammerman, A., & Shafer, G. (2005). Algorithmic learning in a random world. Springer.

[8] Lei, J., G’Sell, M., Rinaldo, A., Tibshirani, R. J., & Wasserman, L. (2018). Distribution-free predictive inference for regression. Journal of the American Statistical Association, 113(523), 1094–1111. https://doi.org/10.1080/01621459.2017.1319823

[9] Bates, S., Angelopoulos, A. N., Lei, L., Malik, J., & Jordan, M. I. (2021). Distribution-free, risk-controlling prediction sets. Journal of the ACM, 68(6), 1–34. https://doi.org/10.1145/3478530

[10] Angelopoulos, A. N., Bates, S., Candès, E. J., Jordan, M. I., & Lei, L. (2022). Learn then test: Calibrating predictive algorithms to achieve risk control. In Proceedings of the 39th International Conference on Machine Learning (Vol. 162, pp. 986–1011). PMLR.

[11] Sadinle, M., Lei, J., & Wasserman, L. (2019). Least ambiguous set-valued classifiers with bounded error levels. Journal of the American Statistical Association, 114(525), 223–234. https://doi.org/10.1080/01621459.2018.1424625

[12] Papadopoulos, H., Proedrou, K., Vovk, V., & Gammerman, A. (2002). Inductive confidence machines for regression. In Proceedings of the European Conference on Machine Learning (pp. 345–356).

[13] Tibshirani, R. J., Foygel Barber, R., Candès, E. J., & Ramdas, A. (2019). Conformal prediction under covariate shift. In Advances in Neural Information Processing Systems (Vol. 32, pp. 2530–2540).

[14] Gibbs, I., & Candès, E. J. (2021). Adaptive conformal inference under distribution shift. In Advances in Neural Information Processing Systems (Vol. 34, pp. 1660–1672).

[15] Foygel Barber, R., Candès, E. J., Ramdas, A., & Tibshirani, R. J. (2023). Conformal prediction beyond exchangeability. The Annals of Statistics, 51(2), 816–845. https://doi.org/10.1214/23-AOS2276

[16] Robins, J. M., Rotnitzky, A., & Zhao, L. P. (1994). Estimation of regression coefficients when some regressors are not always observed. Journal of the American Statistical Association, 89(427), 846–866. https://doi.org/10.1080/01621459.1994.10476818

[17] Gama, J., Žliobaitė, I., Bifet, A., Pechenizkiy, M., & Bouchachia, A. (2014). A survey on concept drift adaptation. ACM Computing Surveys, 46(4), Article 44. https://doi.org/10.1145/2523813

[18] Jiao, Y., Fan, H., Yue, X., Ping, W., Sun, T., & Wang, J. (2026). Dynamic heterogeneous graph contrastive learning for uncovering collusive financial fraud. Scientific Reports. https://doi.org/10.1038/s41598-026-58938-5

[19] Dal Pozzolo, A., Caelen, O., Johnson, R. A., & Bontempi, G. (2015). Calibrating probability with undersampling for unbalanced classification. In Proceedings of the IEEE Symposium Series on Computational Intelligence (pp. 159–166).

[20] Lopez-Rojas, E. A., Elmir, A., & Axelsson, S. (2016). PaySim: A financial mobile money simulator for fraud detection. In Proceedings of the 28th European Modeling & Simulation Symposium (pp. 249–255).

[21] Jesus, S., Pombal, J., Alves, D., Cruz, A., Saleiro, P., Ribeiro, R. P., Gama, J., & Bizarro, P. (2022). Turning the tables: Biased, imbalanced, dynamic tabular datasets for ML evaluation. In Advances in Neural Information Processing Systems (Vol. 35).

[22] Chen, Z., Wang, M., Zeng, Z., & Ping, W. (2026). Uncertainty-aware financial forecasting: Leveraging conformal prediction for risk-adjusted models. IEEE Access, 14, 90053–90077. https://doi.org/10.1109/ACCESS.2026.3702666

[23] Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T.-Y. (2017). LightGBM: A highly efficient gradient boosting decision tree. In Advances in Neural Information Processing Systems (Vol. 30, pp. 3146–3154).

[24] Yue, X., Yang, J., Zhu, R., & Xu, Q. (2026). Adaptive temporal graph neural networks for real-time financial fraud detection under concept drift. Highlights in Business, Economics and Management, 69, 239–250.

[25] Fan, H., Jiao, Y., Wang, M., Chen, J., & Yang, S. (2026). Fraud learns too: Continual graph learning under strategic adversarial drift in dynamic networks. Scientific Reports. https://doi.org/10.1038/s41598-026-60997-7

[26] Dal Pozzolo, A., Caelen, O., Le Borgne, Y.-A., Waterschoot, S., & Bontempi, G. (2014). Learned lessons in credit card fraud detection from a practitioner perspective. Expert Systems with Applications, 41(10), 4915–4928. https://doi.org/10.1016/j.eswa.2014.02.026

[27] Vovk, V. (2012). Conditional validity of inductive conformal predictors. In Proceedings of the Asian Conference on Machine Learning (Vol. 25, pp. 475–490). PMLR.