James Kotary
Welcome to my research homepage. It's a record of the work published throughout my PhD program.
I am interested the integration of constrained optimization and machine learning methods to enable new functionality in AI. Lately, my research has focused on training ML models as fast approximate solvers for hard optimization problems, and on enabling constrained optimization models as components in neural networks to learn data-driven optimal decision making. I'll graduate this year from the University of Virginia, where my advisor is the great Ferdinando Fioretto.
Preprints (Under Review):
Recent Publications:
- Ethan King, James Kotary, Ferdinando Fioretto, Jan Drgona. Metric Learning to Accelerate Convergence of Operator Splitting Methods for Differentiable Parametric Programming. CDC 2024.
- James Kotary, Vincenzo Di Vito, Ferdinando Fioretto, Pascal Van Hentenryck. Learning Joint Models of Prediction and Optimization. ECAI 2024.
- My H Dinh, James Kotary, Ferdinando Fioretto. End-to-End Learning for Fair Multiobjective Optimization Under Uncertainty. UAI 2024.
- My H Dinh, James Kotary, Ferdinando Fioretto. Learning Fair Ranking Policies via Differentiable Optimization of Ordered Weighted Averages. ACM FAccT 2024.
- Jayanta Mandi, James Kotary, Senne Berden, Maxime Mulamba, Victor Bucarey, Tias Guns, Ferdinando Fioretto. Decision-Focused Learning: Foundations, State of the Art, Benchmark and Future Opportunities. JAIR.
- James Kotary, My H Dinh, Ferdinando Fioretto. Backpropagation of Unrolled Solvers with Folded Optimization. IJCAI 2023.
- James Kotary, Vincenzo Di Vito, Ferdinando Fioretto. Differentiable Model Selection for Ensemble Learning. IJCAI 2023.
- James Kotary, Ferdinando Fioretto, Pascal Van Hentenryck, Ziwei Zhu. End-to-End Learning for Fair Ranking Systems. WWW 2022.
- James Kotary, Ferdinando Fioretto, Pascal Van Hentenryck. Fast Approximations for Job Shop Scheduling: A Lagrangian Dual Deep Learning Method. AAAI 2022.
- James Kotary, Ferdinando Fioretto, Pascal Van Hentenryck. Learning Hard Optimization Problems: A Data Generation Perspective. NIPS 2021.
- James Kotary, Ferdinando Fioretto, Pascal Van Hentenryck, Bryan Wilder. End-to-End Constrained Optimization Learning: A Survey. IJCAI 2021 (Survey Track).
Contact:
jk4pn@virginia.edu