Researchers propose WeCon, an efficient Weight-Conditioned neural solver for Multi-Objective Combinatorial Optimization Problems (MOCOPs). It improves weight-conditioned context modeling and preference optimization, addressing limitations of existing methods in weight injection and constructing informative solution pairs for training.
Este trabalho propõe uma formulação rigorosa para a recuperação consciente da diversidade em Geração Aumentada por Recuperação (RAG), abordando a falta de garantias teóricas e escalabilidade dos métodos existentes. A solução utiliza programação quadrática binária com restrição de cardinalidade (CCBQP) e um algoritmo baseado em Frank-Wolfe, demonstrando desempenho superior na fronteira de Pareto de relevância-diversidade e maior velocidade.
This paper addresses Earth Observation satellite scheduling under unknown operational constraints, which must be learned interactively from a binary oracle. The authors introduce Conservative Constraint Acquisition (CCA), a domain-specific procedure, to efficiently identify justified constraints for a simplified model.
This research introduces a composite-move Tabu search (CM-Tabu) algorithm designed for fast and effective spatial redistricting optimization. It tackles the contiguity constraint by expanding the feasible neighborhood to include composite moves, ensuring better exploration and preventing the search from getting trapped in poor local optima.
DynaSchedBench is a new diagnostic framework for the Dynamic Flexible Job Shop Scheduling Problem (DFJSP), addressing limitations in neural combinatorial optimization. It uses a Sequential Event-Space Calibrator (SESC) and Schedule Stress Index (SSI) to rigorously control instance generation and stratify difficulty. This method proves more efficient than evolutionary baselines, enabling rigorous testing of LLM-based scheduling agents.