Skip to content
forked from 4rtemi5/Watten

Reinforcement Learning environment for a local variant of the game Watten including simple evolutionary architecture search.

Notifications You must be signed in to change notification settings

mkofler96/Watten

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Watten

Single notebook Reinforcement Learning for a local variant of the card game "Watten" in its south tyrolean form. It includes a simple evolutionary architecture search and is parallelized quite well.

This was written in a few afternoons after our professor of "Artificial Intelligence" claimed that current machine learning, while being able to learn chess or go cannot learn 'complex' games like watten. I wrote this just to convince myself that Watten is in fact quite easy to learn for a neural network to learn with its frequent and well interpretable rewards and relatively simple rule structure. Even without extensive optimzation or pretraining.
The resulting networks migth not be superhuman at the game but still work really well.

Link to a Colab-Notebook will be provided soon.

Watten (english but not all south tyrolen rules explained): https://en.wikipedia.org/wiki/Watten_(card_game)
Watten (german with pretty much all the rules): https://de.wikipedia.org/wiki/Watten

About

Reinforcement Learning environment for a local variant of the game Watten including simple evolutionary architecture search.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 88.9%
  • Python 11.1%