Inspiration
-Carykh, and Code Bullet
What it does
-It trains an AI to beat a maze based off of the genetic algorithm. the bottom half are killed off and the new ones that replace them are slightly mutated from the survivors
How we built it
- We created a Simple game and then created a neural network tied to a genetic algorithms to evolve it to be more fit and make it to the correct corner of the screen. -We only used python and other default python library's to construct our project ## Future plans
- Right now we only use two layers in the neural network in order to have it be trained enough in time for the submissions, but once the project is over we will revert the algorithm to be 4 layers deep like it was before so it can make much more intelligent solutions.
- we hope that by the end of full training, and with its whole brain unlocked it is able to get to the objective from any spawn point
Log in or sign up for Devpost to join the conversation.