https://github.com/novialriptide/mario-neat

A machine learning genetic algorithm built from the ground up capable of completing Super Mario Bros Level 1-1.
https://github.com/novialriptide/mario-neat

Last synced: 11 days ago
JSON representation

A machine learning genetic algorithm built from the ground up capable of completing Super Mario Bros Level 1-1.

• Host: GitHub
• URL: https://github.com/novialriptide/mario-neat
• Owner: novialriptide
• License: mit
• Created: 2021-04-24T06:40:30.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2023-12-05T07:34:34.000Z (12 months ago)
• Last Synced: 2023-12-05T08:24:33.253Z (12 months ago)
• Language: Lua
• Homepage:
• Size: 4.22 MB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Mario-NEAT

![Video preview](/images/preview.gif)

A machine-learning genetic algorithm built from the ground up using research from the University of Texas-Austin "Neural Evolution of Augmenting Topologies" capable of playing and completing a real Super Mario Bros 1-1 ROM utilizing a Nintendo Entertainment System emulator. This project has been awarded ["Best Technical Skills"](/images/award.png) from the [Department of Computer Science at Rutgers University](https://www.cs.rutgers.edu/) on April 17, 2023.

## Relevant Links

 - [Methodology](http://nn.cs.utexas.edu/downloads/papers/stanley.ec02.pdf), "Neural Evolution of Augmenting Topologies"

 - [FCEUX64](https://fceux.com/web/download.html), the emulator that was used for the project

 - [Super Mario Bros RAM Map](https://datacrystal.romhacking.net/wiki/Super_Mario_Bros.:RAM_map), information regarding how I retrieved data from the NES emulator

 - [YouTube Video Showcase](https://www.youtube.com/shorts/7xQBhGe79_E), a video of me explaining how it works in 1 minute