Inspiration
We wanted to try to learn ML by utilizing this challenge and seeing what it takes to train our own agent.
What it does
It is a ML model that was trained on multiple variations of games in order to decide on what the best decision is on each tick.
How we built it
Python and Pytorch.
Challenges we ran into
Learning how to make the model not be biased towards certain spawn positions in the game, and how to also make it not as deterministic during matches.
Accomplishments that we're proud of
We managed to have the machine learn.
What we learned
We learned machine learning.
What's next for Case Closed
We actually had a lot of fun in seeing the decision making behind the agent. We plan on seeing what other learning methods we can implement into this or see what other ways we can utilize machine learning.
Log in or sign up for Devpost to join the conversation.