First introduction of Reinforcement Learning. It is still very new, I did not focus on performance much, but it is decent. Also, it is faster than simulationist was.
I only focused on training one move (much like simulationist did). Training the whole game, and taking into account current score of agent and opponents score is the next step.
I did not include the Python code used in the training. At some point, I think to even put it in a separate repo.
First introduction of Reinforcement Learning. It is still very new, I did not focus on performance much, but it is decent. Also, it is faster than simulationist was.
I only focused on training one move (much like simulationist did). Training the whole game, and taking into account current score of agent and opponents score is the next step.
I did not include the Python code used in the training. At some point, I think to even put it in a separate repo.