Games are cool, neural networks are cool

How it works

There is a neural network which takes in 9 inputs, shown as the red lines as well as momemtum. Every generation, the network is mutated and tested 4 times. The highest average lives on, until it has evolved into a solution.

Challenges I ran into

Getting the input data was hard, as I had to find a way to pass it into the network real-time

Accomplishments that I'm proud of

I figured out a nice way to display the neural network by using colors for negative/positive, and increasing the width by the weight of each connection.

What I learned

My genetic algorithm was really simple, so it takes a long time to learn even simple things.

What's next for Helicopter Neural Network

Use it for other games

Built With

Share this project: