Skip to content

DeepQ (reinforcement) learns to play Snake with animation

Notifications You must be signed in to change notification settings

wlamnorman/deepQ-learns-snake

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Deep-Q playing Snake

This repository is inspired by Patrick Loeber (https://github.com/patrickloeber) and explores the usage of reinforcement learning applied to the game of Snake.

Ideas for further improvement

Usually the model stops improving around 30-40 score, at this point the snake usually traps itself and has nowhere to go. The reason for this is that the snake is that the agent is not aware of the snakes location except for the head. A potential improvement would therefore be to include the snake's body in state, or, to use a model that allows the agent to have some memory of its previous moves to avoid trapping itself.

About

DeepQ (reinforcement) learns to play Snake with animation

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages