Flounder: An RL Chess Agent

6 January, 2017

For CS221 last quarter, my team wrote a chess engine based on reinforcement learning. I’m posting our paper and poster here for my records. This-playing program isn’t just extraordinarily slow—it’s also absolutely horrible at the game!

See the full report and poster.

Note to self: leave a lot more time at the end to train the model, because it takes a very, very long time.