LandscapeSurrogate: LearningDecisionLossesfor MathematicalOptimizationUnderPartialInformation

Neural Information Processing Systems 

Mathematical optimization problems invarious settings havebeen widely studied, and numerous methods exist to solvethem [25,32].