chairbender / fantasy-football-auction-ai Goto Github PK

Hi @chairbender,

I know that the project is on hold, but being a reinforcement learning practitioner myself, I would like to know what is the current state and performance of the AlphaZero-like implementation.

I had a look at the code in the separate branch, but it is not clear whether the implementation is stable and what its performance is like against the two Q-policy learning methods reported in the README (shallow DQN and conv DQN).

I have previous experience with AlphaZero-like tree search approaches and with different Q-policy learning methods (e.g., Rainbow), so I would be quite interested to know if you had time to compare the two worlds before suspending the project.