Learning to optimize:A tutorial for continuous and mixed-integer optimization  

作　　者：Xiaohan Chen Jialin Liu Wotao Yin 

机构地区：[1]Decision Intelligence Lab,Alibaba DAMO Academy,Bellevue,WA,98004,USA

出　　处：《Science China Mathematics》2024年第6期1191-1262,共72页中国科学（数学）（英文版）

摘　　要：Learning to optimize(L2O)stands at the intersection of traditional optimization and machine learning,utilizing the capabilities of machine learning to enhance conventional optimization techniques.As real-world optimization problems frequently share common structures,L2O provides a tool to exploit these structures for better or faster solutions.This tutorial dives deep into L2O techniques,introducing how to accelerate optimization algorithms,promptly estimate the solutions,or even reshape the optimization problem itself,making it more adaptive to real-world applications.By considering the prerequisites for successful applications of L2O and the structure of the optimization problems at hand,this tutorial provides a comprehensive guide for practitioners and researchers alike.

关 键 词：AI for mathematics(AI4Math) learning to optimize algorithm unrolling plug-and-play methods differentiable programming machine learning for combinatorial optimization(ML4CO) 

分 类 号：O224[理学—运筹学与控制论] TP181[理学—数学]