aalto1 untyped-item.component.html

Computing approximate Nash equilibria in action graph games with iterated polymatrix approximation

Loading...
Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

Perustieteiden korkeakoulu | Bachelor's thesis

Department

Mcode

SCI3029

Language

en

Pages

17+1

Series

Abstract

This thesis explores computing approximate Nash equilibria in Action Graph Games (AGGs) using the Iterated Polymatrix Approximation (IPA) algorithm. AGGs are a graph-based compact representation of multiplayer games where the interactions between players are captured through a graph structure, offering an efficient framework for analyzing strategic behavior numerically. The IPA algorithm iteratively refines approximations of Nash equilibria by representing complex multiplayer games locally as multiple two-player games. This work implements the IPA algorithm on action graph form games and evaluates its efficiency and accuracy with comparisons to the original implementation. The results show that while the accuracy does not increase with this new approach, the speedup and reduced complexity in scaling games to more players are significant.

Description

Supervisor

Salo, Ahti

Thesis advisor

Salo, Ahti

Other note

Citation

Endorsement

Review

Supplemented By

Referenced By