Seth Bling's Mar/IO project
What it does
This algorithm could be applied to many different problems, but currently I'm training it to play a simple game. It's task is to try and make a character avoid projectiles coming from all angles
How I built it
This project was built using java and a graphics library I wrote previously.
Challenges I ran into
This algorithm has many small parts that need to be tweaked to make sure evolution runs smoothly. There are many nobs and dials that need to be tuned and we simply didn't have the time to test everything. The network also takes hours to train fully so we were only able to run one full session.
Accomplishments that I'm proud of
The training did work. Although it currently cant be a human, it is able to learn and play the game.
What I learned
We learned a lot about machine learning and the basic principles of neural networks.
What's next for Java NEAT algorithm Implementation
We want apply it to more then just a simple game I created. Next we would like to apply it to an emulated NES game.