Implementing Neural Networks in Standard's AI
The plan is to use neural nets to train the AI for Standard.
I have the learning code worked out, the remaining work involves writing some functions to let the learner interface with the game.
Steps
-Figure out learning code (Completed)
-Build functions to let the learning code interface with the game (in progress)
-Create an alternate map where turns and movement are faster to allow for more iterations in a shorter amount of time (Next)
-Work out a way to retain the learning across multiple maps and instances (Incomplete)
More detailed updates to follow.