If you don't have prior experience with neural networks:
- Learn however would make the most sense for you. There are great introductory resources here and here
- Move on to the next bullet
If you're already comfortable with neural networks:
- Spend some time become familiar with Reinforcement Learning. You can use the resources here and here
- Find me on the Hack Club Slack (@Vihaan Sondhi) and join #reinforced (linked at the bottom) if you have any questions or need help understanding something. We'll be working closely one on one throughout the process anyway so don't be shy! It's better to be safe than waste hours later.
Once you're done with all of this, contact me using the methods above and I'll send you a brief quiz (no pressure!)
- Send me your game idea for approval. I'll help you brainstorm the structure of the game so that it's a good environment for RL down the road.
- The game can be an entirely new game of your creation or a version of an existing game (wordle, snake, pong, etc).
- Build the game using vanilla Python, Pygame, or Unity. If you want to have both a clean-looking game and a relatively easy approach to solving your game with RL, I highly recommend the last.