Help
Cart
Skip main navigation

Decision-Focused Learning: When and Why Traditional Prediction Models Fail

Published Online:1 Nov 2026https://doi.org/10.1287/educ.2026.0303
Acknowledgments

M. Liu gratefully acknowledges Erick Delage, Tito Homem-de-Mello, and Vishal Gupta for valuable feedback and the editors and four anonymous reviewers for helpful comments and suggestions.

References

  • [1] A. Agrawal, B. Amos, S. Barratt, S. Boyd, S. Diamond, and J. Z. Kolter. Differentiable convex optimization layers. H. Wallach, H. Larochelle, A. Beygelzimer, F. d’Alché-Buc, E. Fox, and R. Garnett, eds. Advances in Neural Information Processing Systems, vol. 32. Curran Associates Inc., Red Hook, NY, 9558–9570, 2019.Google Scholar
  • [2] B. Amos and J. Z. Kolter. OptNet: Differentiable optimization as a layer in neural networks. D. Precup and Y. W. Teh, eds. Proceedings of the International Conference on Machine Learning,vol. 70. JMLR.org, 136–145, 2017.Google Scholar
  • [3] G. Y. Ban and C. Rudin. The big data newsvendor: Practical insights from machine learning. Operations Research 67(1):90–108, 2019.Link, Google Scholar
  • [4] O. Bennouna, A. Bennouna, S. Amin, and A. Ozdaglar. What data enables optimal decisions? An exact characterization for linear optimization. Preprint, submitted May 27, https://arxiv.org/abs/2505.21692, 2025.Google Scholar
  • [5] S. Berden, A. İ. Mahmutoğulları, D. Tsouros, and T. Guns. Solver-free decision-focused learning for linear optimization problems. Preprint, submitted May 28, https://arxiv.org/abs/2505.22224, 2025.Google Scholar
  • [6] D. Bertsimas and N. Kallus. From predictive to prescriptive analytics. Management Science 66(3):1025–1044, 2020.Link, Google Scholar
  • [7] D. Bertsimas and N. Mundru. Optimization-based scenario reduction for data-driven two-stage stochastic optimization. Operations Research 71(4):1343–1361, 2023.Link, Google Scholar
  • [8] V. Bucarey, S. Calderón, G. Muñoz, and F. Semet. Decision-focused predictions via pessimistic bilevel optimization: A computational study. B. Dilkina, ed. Integration of Constraint Programming, Artificial Intelligence, and Operations Research: 21st International Conference, CPAIOR 2024, Proceedings, Part I, Lecture Notes in Computer Science 14742. Springer Nature Switzerland, Cham, 127–135, 2024.Google Scholar
  • [9] A. Capitaine, M. Haddouche, E. Moulines, M. I. Jordan, E. Boursier, and A. Durmus. Online decision-focused learning. Proceedings of the International Conference on Learning Representations, 2026.Google Scholar
  • [10] X. Cheng, Y. Zhu, and Y. Xie. Generative models for decision-making under distributional shift. Preprint, submitted April 6, https://arxiv.org/abs/2604.04342, 2026.Google Scholar
  • [11] V. Chernozhukov, C. Hansen, N. Kallus, M. Spindler, and V. Syrgkanis. Applied causal inference powered by ML and AI. Preprint, submitted March 4, https://arxiv.org/abs/2403.02467, 2024.Google Scholar
  • [12] L. Y. Chu, J. G. Shanthikumar, and Z. J. M. Shen. Solving operational statistics via a Bayesian analysis. Operations Research Letters 36(1):110–116, 2008.Crossref, Google Scholar
  • [13] A. T. H. Chung, J. Abdulai, P. Bayoh, L. Sandi, F. Smart, H. Bastani, and O. Bastani. Improving access to essential medicines via decision-aware machine learning. Nature 653(8116):1178–1183, 2026.Google Scholar
  • [14] R. Cristian, P. Harsha, G. Perakis, and B. Quanz. Efficient end-to-end learning for decision-making: A meta-optimization approach. Preprint, submitted May 16, https://arxiv.org/abs/2505.11360, 2025.Google Scholar
  • [15] P. L. Donti, J. Z. Kolter, and B. Amos. Task-based end-to-end model learning in stochastic optimization. I. Guyon, U. von Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, eds. Advances in Neural Information Processing Systems, vol. 30. Curran Associates Inc., Red Hook, NY, 5484–5494, 2017.Google Scholar
  • [16] O. El Balghiti, A. N. Elmachtoub, P. Grigas, and A. Tewari. Generalization bounds in the predict-then-optimize framework. Mathematics of Operations Research 48(4):2043–2065, 2023.Link, Google Scholar
  • [17] A. N. Elmachtoub and P. Grigas. Smart “predict, then optimize”. Management Science 68(1):9–26, 2022.Link, Google Scholar
  • [18] A. N. Elmachtoub, H. Lam, H. Lan, and H. Zhang. Dissecting the impact of model misspecification in data-driven optimization. Preprint, submitted March 1, https://arxiv.org/abs/2503.00626, 2025.Google Scholar
  • [19] A. N. Elmachtoub, H. Lam, H. Zhang, and Y. Zhao. Estimate-then-optimize versus integrated-estimation-optimization versus sample average approximation: A stochastic dominance perspective. Preprint, submitted April 13, https://arxiv.org/abs/2304.06833, 2023.Google Scholar
  • [20] C. Er and M. Liu. Decision-focused bias correction for fluid approximation. Preprint, submitted December 4, https://arxiv.org/abs/2512.15726, 2025.Google Scholar
  • [21] Q. Feng, J. G. Shanthikumar, and J. Wu. Contextual data-integrated newsvendor solution with operational data analytics (ODA). Management Science 71(11):9384–9403, 2025.Link, Google Scholar
  • [22] V. Gupta. End-to-end learning and optimization: Course reader. Course reader, USC Marshall School of Business. Accessed May 31, 2026, https://www.dropbox.com/scl/fo/hku67dioasy5rz0pxgs08/AOl182gWOvBUvDujNd8m8f4?dl=0&e=1&rlkey=v0wumq8vbwqe0dia5dk6acw37, 2026.Google Scholar
  • [23] N. Ho-Nguyen and F. Kılınç-Karzan. Risk guarantees for end-to-end prediction and optimization processes. Management Science 68(12):8680–8698, 2022.Link, Google Scholar
  • [24] T. Homem-de Mello, J. Valencia, F. Lagos, and G. Lagos. Forecasting outside the box: Application-driven optimal pointwise forecasts for stochastic optimization. Preprint, submitted November 5, https://arxiv.org/abs/2411.03520, 2024.Google Scholar
  • [25] Y. Hu, N. Kallus, and X. Mao. Fast rates for contextual linear optimization. Management Science 68(6):4236–4245, 2022.Link, Google Scholar
  • [26] X. Hu, J. Lee, and J. Lee. Two-stage predict+ optimize for MILPs with unknown parameters in constraints. A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S. Levine, eds. Advances in Neural Information Processing Systems, vol. 36. Curran Associates Inc., Red Hook, NY, 14247–14272, 2023.Google Scholar
  • [27] Y. Hu, N. Kallus, X. Mao, and Y. Wu. Contextual linear optimization under partial feedback. Preprint, submitted November 15, https://doi.org/10.2139/ssrn.5724783, 2025.Google Scholar
  • [28] M. Huang and V. Gupta. Decision-focused learning with directional gradients. A. Globerson, L. Mackey, D. Belgrave, A. Fan, U. Paquet, J. Tomczak, and C. Zhang, eds. Advances in Neural Information Processing Systems, vol. 37. Curran Associates Inc., Red Hook, NY, 79194–79220, 2024.Google Scholar
  • [29] H. Im, W. Benslimane, and P. Grigas. Smart surrogate losses for contextual stochastic linear optimization with robust constraints. Preprint, submitted May 28, https://arxiv.org/abs/2505.22881, 2025.Google Scholar
  • [30] J. Kotary, V. Di Vito, J. Christopher, P. Van Hentenryck, and F. Fioretto. Predict-then-optimize by proxy: Learning joint models of prediction and optimization. Preprint, submitted November 22, https://arxiv.org/abs/2311.13087, 2023.Google Scholar
  • [31] H. Lan, L. Liao, A. N. Elmachtoub, C. Kroer, H. Lam, and H. Zhang. The bias-variance tradeoff in data-driven optimization: A local misspecification perspective. Preprint, submitted October 21, https://arxiv.org/abs/2510.18215, 2025.Google Scholar
  • [32] J. Lee, S. Jin, and Y. Lee. Decision-focused learning via tangent-space projection of prediction error. A. Agarwal, M. Dudík, M. Jaggi, and S. Li, eds. Proceedings of the 43rd International Conference on Machine Learning, Proceedings of Machine Learning Research 306. PMLR, New York, 2026.Google Scholar
  • [33] H. Liu and P. Grigas. Online contextual decision-making with a smart predict-then-optimize method. Preprint, submitted June 15, https://arxiv.org/abs/2206.07316, 2022.Google Scholar
  • [34] S. Liu and M. Liu. Decision-focused optimal transport. Preprint, submitted February 2, https://arxiv.org/abs/2602.02800, 2026.Google Scholar
  • [35] M. Liu, J. Cao, and Z. J. M. Shen. Value of one data point: Active label acquisition in assortment optimization. Preprint, submitted June 26, https://doi.org/10.2139/ssrn.4487888, 2023.Google Scholar
  • [36] M. Liu, Y. Bai, M. Qi, and Z. J. M. Shen. Inventory management with transformer: Automated decision making for order timing and quantity. Service Science, ePub ahead of print April 7, https://doi.org/10.1287/serv.2024.0236, 2026.Google Scholar
  • [37] M. Liu, P. Grigas, H. Liu, and Z. J. M. Shen. Active learning in the predict-then-optimize framework: A margin-based approach. Preprint, submitted May 11, https://arxiv.org/abs/2305.06584, 2023.Google Scholar
  • [38] J. Mandi, J. Kotary, S. Berden, M. Mulamba, V. Bucarey, T. Guns, and F. Fioretto. Decision-focused learning: Foundations, state of the art, benchmark and future opportunities. Journal of Artificial Intelligence Research 81:1623–1701, 2024.Crossref, Google Scholar
  • [39] M. Qi, P. Grigas, and Z. J. Shen. Integrated conditional estimation-optimization. Operations Research 74(3):1604–1625, 2025.Google Scholar
  • [40] M. Qi, Y. Shi, Y. Qi, C. Ma, R. Yuan, D. Wu, and Z. J. Shen. A practical end-to-end inventory management model with deep learning. Management Science 69(2):759–773, 2023.Link, Google Scholar
  • [41] P. Rodriguez-Diaz, L. Kong, K. Wang, D. Alvarez-Melis, and M. Tambe. What is the right notion of distance between predict-then-optimize tasks? Preprint, submitted September 11, https://arxiv.org/abs/2409.06997, 2024.Google Scholar
  • [42] U. Sadana, A. Chenreddy, E. Delage, A. Forel, E. Frejinger, and T. Vidal. A survey of contextual optimization methods for decision-making under uncertainty. European Journal of Operational Research 320(2):271–289, 2025.Crossref, Google Scholar
  • [43] P. J. Schneider and D. Kuhn. Soft-radial projection for constrained end-to-end learning. Preprint, submitted February 3, https://arxiv.org/abs/2602.03461, 2026.Google Scholar
  • [44] S. Shah, B. Wilder, A. Perrault, and M. Tambe. Leaving the nest: Going beyond local loss functions for predict-then-optimize. Proceedings of the AAAI Conference on Artificial Intelligence, vol. 38, 14902–14909, 2024.Google Scholar
  • [45] S. Shah, K. Wang, B. Wilder, A. Perrault, and M. Tambe. Decision-focused learning without decision-making: Learning locally optimized decision losses. S. Koyejo, S. Mohamed, A. Agarwal, D. Belgrave, K. Cho, and A. Oh, eds. Advances in Neural Information Processing Systems, vol. 35. Curran Associates Inc., Red Hook, NY, 1320–1332, 2022.Google Scholar
  • [46] B. Tang and E. B. Khalil. PyEPO: A PyTorch-based end-to-end predict-then-optimize library for linear and integer programming. Mathematical Programming Computation 16:297–335, 2024.Google Scholar
  • [47] B. Wan and M. Liu. A solver-free training method for predict-then-optimize. A. Agarwal, M. Dudík, M. Jaggi, and S. Li, eds. Proceedings of the 43rd International Conference on Machine Learning, Proceedings of Machine Learning Research 306. PMLR, New York, 2026. Google Scholar
  • [48] B. Wan, M. Liu, P. Grigas, and Z. J. M. Shen. Decision-focused sequential experimental design: A directional uncertainty-guided approach. Preprint, submitted February 5, https://arxiv.org/abs/2602.05340, 2026.Google Scholar
  • [49] K. Wang, B. Wilder, A. Perrault, and M. Tambe. Automatically learning compact quality-aware surrogates for optimization problems. H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, eds. Advances in Neural Information Processing Systems, vol. 33. Curran Associates Inc., Red Hook, NY, 9586–9596, 2020.Google Scholar
  • [50] B. Wilder, E. Ewing, B. Dilkina, and M. Tambe. End to end learning and optimization on graphs. H. Wallach, H. Larochelle, A. Beygelzimer, F. d’Alché-Buc, E. Fox, and R. Garnett, eds. Advances in Neural Information Processing Systems, vol. 32. Curran Associates Inc., Red Hook, NY, 4672–4683, 2019.Google Scholar
  • [51] C. Yeh, N. Christianson, A. Wierman, and Y. Yue. Conformal risk training: End-to-end optimization of conformal risk control. Preprint, submitted October 9, https://arxiv.org/abs/2510.08748, 2025.Google Scholar
  • [52] J. Zhao. Experimental design for causal inference through an optimization lens. V. Babich, P. Dewan, and J. Pender, eds. INFORMS TutORials in Operations, Catonsville, MD, 146–188, 2024.Link, Google Scholar
Previous
Back to Top
Next
INFORMS site uses cookies to store information on your computer. Some are essential to make our site work; Others help us improve the user experience. By using this site, you consent to the placement of these cookies. Please read our Privacy Statement to learn more.
Agree