Geodesic
Linear and superlinear convergence of a potential reduction algorithm for generalized Nash equilibrium problems
Minimax theory and its applications, Tome 9 (2024) no. 1
Cet article a éte moissonné depuis la source Minimax Theory and its Applications website

Voir la notice de l'article

Wepresent a detailed convergence analysis of the potential reduction algorithm for generalized Nash equilibrium problems (GNEPs), that is known to be a robust method for solving those problems. Weprove Q-linear convergence of the merit function and R-linear convergence of the distance of the iterates to the set of KKT-points of the GNEP. Furthermore, we show that the stepsize is bounded from below, implying finite termination of the method for prescribed accuracy. Using a non-fixed linesearch parameter we prove superlinear convergence. Further, we give additional assumptions to transfer the convergence rates to an inexact potential reduction algorithm. By our analysis, we also discover indicators that could be used to estimate the active set at an accumulation point, and hence at a generalized Nash equilibrium, which might be exploited numerically.
Mots-clés : Generalized Nash equilibrium problem, potential reduction algorithm, linear conver gence, superlinear convergence 
@article{MTA_2024_9_1_a3,
     author = {Axel Dreves},
     title = {Linear and superlinear convergence of a potential reduction algorithm for generalized {Nash} equilibrium problems},
     journal = {Minimax theory and its applications},
     year = {2024},
     volume = {9},
     number = {1},
     zbl = {1544.90185},
     url = {http://geodesic.mathdoc.fr/item/MTA_2024_9_1_a3/}
}
TY  - JOUR
AU  - Axel Dreves
TI  - Linear and superlinear convergence of a potential reduction algorithm for generalized Nash equilibrium problems
JO  - Minimax theory and its applications
PY  - 2024
VL  - 9
IS  - 1
UR  - http://geodesic.mathdoc.fr/item/MTA_2024_9_1_a3/
ID  - MTA_2024_9_1_a3
ER  - 
%0 Journal Article
%A Axel Dreves
%T Linear and superlinear convergence of a potential reduction algorithm for generalized Nash equilibrium problems
%J Minimax theory and its applications
%D 2024
%V 9
%N 1
%U http://geodesic.mathdoc.fr/item/MTA_2024_9_1_a3/
%F MTA_2024_9_1_a3
Axel Dreves. Linear and superlinear convergence of a potential reduction algorithm for generalized Nash equilibrium problems. Minimax theory and its applications, Tome 9 (2024) no. 1. http://geodesic.mathdoc.fr/item/MTA_2024_9_1_a3/