Abstract
An algorithm is proposed for computing an unconstrained minimax, based on differential equations with suitable stabilization terms. Methods for accelerating the convergence are discussed. For computing a constrained minimax, the augmented Lagrangian algorithm of Powell, Hestenes and Rockafellar is generalized to minimax, assuming the unconstrained minimax algorithm as a subroutine. An estimate of the convergence rate is obtained.
Similar content being viewed by others
References
Auslender A (1972) Problèmes de minimax via l'analyse convexe et les inégalités variationelles: théorie et algorithms. Lecture Notes in Economies and Math. Systems 77. Springer, Berlin
Bellman R (1953) Stability theory of differential equations. Dover, New York
Bertsekas DP (1982) Constrained optimization and Lagrange multipliers. Academic Press, New York
Dem'janov VF (1967) Successive approximations for finding saddle points. Soviet Math Dokl8(6): 1350–1353
Goh CJ, Teo KL (1987) MISER, an optimal control software. Dept of Industrial Engineering. National University of Singapore
Hartung J (1982) An extension of Sion's minimax theorem with an application to a method for constrained games. Pac J Math 103:401–408
Shimizu KA, Amoshi E (1980) Necessary condition for min-max problems and algorithms by a relaxation procedure. IEEE Trans on Automatic Control, AC-25:62–66
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Craven, B.D. An algorithm for minimax. ZOR - Methods and Models of Operations Research 35, 425–434 (1991). https://doi.org/10.1007/BF01498333
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF01498333