https://github.com/desophos/gamer-evolution

A framework for evolving game-theoretic agents.
https://github.com/desophos/gamer-evolution

evolution finite-state-machine game-theory genetic-algorithm

Last synced: 18 days ago
JSON representation

A framework for evolving game-theoretic agents.

• Host: GitHub
• URL: https://github.com/desophos/gamer-evolution
• Owner: desophos
• License: mit
• Created: 2021-02-03T06:03:33.000Z (about 4 years ago)
• Default Branch: main
• Last Pushed: 2022-09-07T07:34:05.000Z (over 2 years ago)
• Last Synced: 2024-11-29T03:29:10.709Z (3 months ago)
• Topics: evolution, finite-state-machine, game-theory, genetic-algorithm
• Language: Haskell
• Homepage:
• Size: 154 KB
• Stars: 1
• Watchers: 2
• Forks: 0
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • Changelog: ChangeLog.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Gamer Evolution

## Overview

gamer-evolution is a framework for evolving game-theoretic agents. The agents are represented by finite state machines and evolved with a genetic algorithm in which the fitness function is the game being played. It is designed to support any game, although it's currently limited to 2-player games. `app/Main.hs` includes an example usage: graphing a population's average fitness over generations of evolution.

## Background

This is a rewrite from scratch of my old project [winning-the-ipd](https://github.com/desophos/winning-the-ipd). I decided to rebuild it because it was broken, the code was poorly architected, and it was a good excuse to learn Haskell.