Reinforcement Nash Equilibrium Solver
Nash Equilibrium (NE) is the canonical solution concept of game theory, which provides an elegant tool to understand the rationalities. Computing NE in two- or multi-player general-sum games is PPAD-Complete. Therefore, in this work, we propose REinforcement Nash Equilibrium Solver (RENES), which tr...
Saved in:
Main Authors: | , , , , , , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2024
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/9044 https://ink.library.smu.edu.sg/context/sis_research/article/10047/viewcontent/ReinforcementNash_pvoa_cc_by.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Singapore Management University |
Language: | English |
id |
sg-smu-ink.sis_research-10047 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-100472024-07-25T07:53:03Z Reinforcement Nash Equilibrium Solver WANG, Xinrun YANG, Chang LI, Shuxin LI, Pengdeng HUANG, Xiao CHAN, Hau AN, Bo Nash Equilibrium (NE) is the canonical solution concept of game theory, which provides an elegant tool to understand the rationalities. Computing NE in two- or multi-player general-sum games is PPAD-Complete. Therefore, in this work, we propose REinforcement Nash Equilibrium Solver (RENES), which trains a single policy to modify the games with different sizes and applies the solvers on the modified games where the obtained solution is evaluated on the original games. Specifically, our contributions are threefold. i) We represent the games as ��-rank response graphs and leverage graph neural network (GNN) to handle the games with different sizes as inputs; ii) We use tensor decomposition, e.g., canonical polyadic (CP), to make the dimension of modifying actions fixed for games with different sizes; iii) We train the modifying strategy for games with the widely-used proximal policy optimization (PPO) and apply the solvers to solve the modified games, where the obtained solution is evaluated on original games. Extensive experiments on large-scale normal-form games show that our method can further improve the approximation of NE of different solvers, i.e., ��-rank, CE, FP and PRD, and can be generalized to unseen games. 2024-05-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/9044 https://ink.library.smu.edu.sg/context/sis_research/article/10047/viewcontent/ReinforcementNash_pvoa_cc_by.pdf http://creativecommons.org/licenses/by/3.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Game Theory Generalizability Reinforcement Learning Artificial Intelligence and Robotics Theory and Algorithms |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
Game Theory Generalizability Reinforcement Learning Artificial Intelligence and Robotics Theory and Algorithms |
spellingShingle |
Game Theory Generalizability Reinforcement Learning Artificial Intelligence and Robotics Theory and Algorithms WANG, Xinrun YANG, Chang LI, Shuxin LI, Pengdeng HUANG, Xiao CHAN, Hau AN, Bo Reinforcement Nash Equilibrium Solver |
description |
Nash Equilibrium (NE) is the canonical solution concept of game theory, which provides an elegant tool to understand the rationalities. Computing NE in two- or multi-player general-sum games is PPAD-Complete. Therefore, in this work, we propose REinforcement Nash Equilibrium Solver (RENES), which trains a single policy to modify the games with different sizes and applies the solvers on the modified games where the obtained solution is evaluated on the original games. Specifically, our contributions are threefold. i) We represent the games as ��-rank response graphs and leverage graph neural network (GNN) to handle the games with different sizes as inputs; ii) We use tensor decomposition, e.g., canonical polyadic (CP), to make the dimension of modifying actions fixed for games with different sizes; iii) We train the modifying strategy for games with the widely-used proximal policy optimization (PPO) and apply the solvers to solve the modified games, where the obtained solution is evaluated on original games. Extensive experiments on large-scale normal-form games show that our method can further improve the approximation of NE of different solvers, i.e., ��-rank, CE, FP and PRD, and can be generalized to unseen games. |
format |
text |
author |
WANG, Xinrun YANG, Chang LI, Shuxin LI, Pengdeng HUANG, Xiao CHAN, Hau AN, Bo |
author_facet |
WANG, Xinrun YANG, Chang LI, Shuxin LI, Pengdeng HUANG, Xiao CHAN, Hau AN, Bo |
author_sort |
WANG, Xinrun |
title |
Reinforcement Nash Equilibrium Solver |
title_short |
Reinforcement Nash Equilibrium Solver |
title_full |
Reinforcement Nash Equilibrium Solver |
title_fullStr |
Reinforcement Nash Equilibrium Solver |
title_full_unstemmed |
Reinforcement Nash Equilibrium Solver |
title_sort |
reinforcement nash equilibrium solver |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2024 |
url |
https://ink.library.smu.edu.sg/sis_research/9044 https://ink.library.smu.edu.sg/context/sis_research/article/10047/viewcontent/ReinforcementNash_pvoa_cc_by.pdf |
_version_ |
1814047715820044288 |