PhD Candidate, Industrial Engineering and Operations Research · Penn State
AI-integrated discrete optimization at scale
Many planning problems pair a decision that is discrete with a cost that is only known through data. The usual response is to call the learned model as a black box inside a search, which gives the optimizer no structure to reason with. I take the other route and represent the trained network exactly inside a mixed-integer program, so the cost becomes part of the model the solver sees rather than an oracle it queries.
Doing this honestly makes the formulation large, so the work runs on to the methods that keep it solvable. That means decomposition, penalty methods for optimizing over deep rectified linear networks, and parallel implementations that reach instance sizes a single solver call cannot. The line began with integrated location and routing, where the routing cost of a cluster of customers is learned once and then reused inside the siting decision.
Publications
W. Kaleem, D. Lee, C. Kwon, A. Subramanyam.
“Neural embedded mixed-integer optimization for location-routing problems.”
Under review at Transportation Research Part E, 2026.
W. Kaleem, D. Lee, C. Kwon, A. Subramanyam.
“Neural embedded mixed-integer optimization for location-routing problems.”
NeurIPS Workshop on Differentiable Learning of Combinatorial Algorithms, 2025.
[openreview]
W. Kaleem, H. Ayala, A. Subramanyam.
“Neural embedded optimization for integrated location and routing problems.”
IISE Annual Conference Proceedings, 1–6, 2024.
Best Paper Award, IISE Logistics and Supply Chain Division.
W. Kaleem, D. Lee, C. Kwon, A. Subramanyam.
NEO-LRP, neural embedded optimization for location-routing problems. Software.
[github]
Talks
Neural embedded mixed-integer optimization for location-routing problems.
INFORMS TSL Conference, MIT, Cambridge, MA. July 2026.INFORMS Computing Society Conference, Toronto. March 2025.
Optimization over deep ReLU networks using penalty methods.
INFORMS Annual Meeting, Atlanta, GA. October 2025.
Neural embedded optimization for sets.
IISE Annual Conference, Montreal. May 2024.INFORMS Optimization Society Conference, Houston, TX. March 2024.
Neural embedded optimization for integrated location and routing problems.
IISE Annual Conference, Montreal. May 2024.
Offline model-based routing optimization with MIP-representable deep set surrogates.
Poster, Mixed Integer Programming Workshop, Stamford, CT. May 2026.