Skip to content

Code built for the testing and analysis of mixed optimization criteria in multi-objective multi-agent reinforcement learning

License

Notifications You must be signed in to change notification settings

yaserBK/equilibria_monfg

 
 

Repository files navigation

An analysis of Nash Equilibrium under the use of mixed optimization criteria in multi-objective normal-form games

An analysis of Nash Equilibrium in under a mix of SER (scalarised expected returns) and ESR (expected scalarized returns) multi-objective optimization criteria

Paper citation

This fork expands upon the code used for experimentation in the following paper:

@article{radulescu2020equilibria,
author="R\u{a}dulescu, Roxana and Mannion, Patrick and Zhang, Yijie and Roijers, Diederik M. and Now{\'e}, Ann",
title="A utility-based analysis of equilibria in multi-objective normal form games",
journal="Knowledge Engineering Review",
year="2020",
note="In press"
}

Getting Started

The code base contains 3 types of Q-learning agents

  1. SER
  2. ESR
  3. ESR with Opponent Modelling capabilities

Experiments can be run using command line arguments such as

python MONFG.py -row ESR -column SER -game game1 -runs 10

or

python MONFG.py -row ESROppo -column SER

The code-base can also be slightly adjusted as needed to very easily run single-objective normal form games such as chicken, battle of the sexes, the prisoners dilemma, etc.

License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details

About

Code built for the testing and analysis of mixed optimization criteria in multi-objective multi-agent reinforcement learning

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%