Zhiguang Cao

AutoSND uncovers more effective and interpretable network dismantling heuristics by transforming execution evidence into actionable structural policies.

Using a large language model as an external trainer, LaT boosts multi-task neural solvers' performance on diverse Vehicle Routing Problems without the computational burden of traditional meta-learning.

AGDN not only solves the Traveling Salesman Problem more effectively but also maintains optimal connectivity in solutions that traditional methods often overlook.

VaFM outperforms traditional methods by effectively integrating visual semantics into vehicle routing, addressing complex constraints that were previously overlooked.

By explicitly modeling operation sequences on machines with a novel heterogeneous disjunctive graph, MIStar significantly boosts the performance of improvement-based methods for flexible job-shop scheduling.

Achieve state-of-the-art results on multi-task vehicle routing problems by dynamically adapting to evolving constraints, even when those constraints are unseen during training.

Neural routing solvers can now efficiently tackle hard constraints thanks to Construct-and-Refine (CaR), which slashes the refinement steps needed by 500x while boosting solution quality.