Good dopamine article

Very complete, and reminds me that I need to refresh my memory on temporal difference learning, with maybe an implementation of something in order.

… like, say a learning Pente implementation based on TD-Gammon? Didn’t I start to do that once, and then get distracted? I believe I did.

… or wait, heck, TD takes a state and generates a prediction error; why not use it for robot training? Why not indeed!

Advertisements

Post a Comment

Required fields are marked *
*
*

%d bloggers like this: