Skip to content

A simple simulation of the neuroevolution of a program learning how to play the game 'Flappy Bird'.

Notifications You must be signed in to change notification settings

arnavkohli/flappy-bird-neuroevolution

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Flappy Bird Neuroevolution

The program learns how to play the game, through the process of neuro-evolution.

Parameters:

  • Population count in each generation
  • Speed of the Pipes
  • Length of gaps of the pipes
  • Distance between consecutive pipes
  • the Speed of evolution
The number of generations it can take, to get the 'perfect' bird can vary as they depend upon the above parameters.

Basically, the easier the game and more the population count per generation, the faster the process.
User can make the game more difficult by increasing the speed of the pipes, decreasing the lengths of gaps and the distance between consecutive pipes.

Speed of evolution just means how sped up this whole process is, you can increase it to get faster results, but it won't effect the number of generations it takes to get to the 'perfect' bird.

Increase the speed of evolution initially, and slow it down after a few generations. You'll see that the birds are a bit more 'smarter' than what they were a few generations ago.

Keep it running and eventually you will come across a somewhat perfect bird, if not perfect.

Inspired by the same project Daniel Shiffman did as a coding challenge on his Youtube channel ('The Coding Train').

About

A simple simulation of the neuroevolution of a program learning how to play the game 'Flappy Bird'.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages