Conservative Parametric Optimality and the Ridge Method for Tame Min-Max Problems

被引:1
作者
Pauwels, Edouard [1 ]
机构
[1] Univ Toulouse, CNRS, Inst Univ France IUF, IRIT, Toulouse, France
关键词
Min-max problems; Ridge algorithm; Parametric optimality; Conservative gradients; Definable sets; O-minimal structures; Clarke subdifferential; First order methods; NONCONVEX;
D O I
10.1007/s11228-023-00682-3
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
We study the ridge method for min-max problems, and investigate its convergence without any convexity, differentiability or qualification assumption. The central issue is to determine whether the "parametric optimality formula" provides a conservative gradient, a notion of generalized derivative well suited for optimization. The answer to this question is positive in a semi-algebraic, and more generally definable, context. As a consequence, the ridge method applied to definable objectives is proved to have a minimizing behavior and to converge to a set of equilibria which satisfy an optimality condition. Definability is key to our proof: we show that for a more general class of nonsmooth functions, conservativity of the parametric optimality formula may fail, resulting in an absurd behavior of the ridge method.
引用
收藏
页数:24
相关论文
共 46 条
[1]  
Ablin P., 2020, Super-efficiency of automatic differentiation for functions defined as a minimum
[2]  
Agrawal A, 2019, ADV NEUR IN, V32
[3]  
Aliprantis C.D., 1999, INFINITE DIMENSIONAL, V2, DOI [10.1007/978-3-662-03961-8, DOI 10.1007/978-3-662-03961-8]
[4]  
Ambrosio L, 2008, LECT MATH, P1
[5]  
Amos B, 2017, PR MACH LEARN RES, V70
[6]  
Arjovsky Chintala, 2017, INT C MACHINE LEARNI
[7]   Convergence of descent methods for semi-algebraic and tame problems: proximal algorithms, forward-backward splitting, and regularized Gauss-Seidel methods [J].
Attouch, Hedy ;
Bolte, Jerome ;
Svaiter, Benar Fux .
MATHEMATICAL PROGRAMMING, 2013, 137 (1-2) :91-129
[8]   Stochastic approximations and differential inclusions [J].
Benaïm, M ;
Hofbauer, J ;
Sorin, S .
SIAM JOURNAL ON CONTROL AND OPTIMIZATION, 2005, 44 (01) :328-348
[9]  
Berthet Quentin., 2020, Advances in Neural Information Processing Sys-tems, V33, P9508
[10]  
Bertsekas D. P., 1971, PhD thesis