Alpha Go Zero: How and Why It Works

Re:Alpha Go Zero: How and Why It Works by nilkn on Sunday November 05, 2017 @11:37AM (#15630129)

Are you sure AlphaGo Zero even uses temporal difference learning? The Nature paper suggests it does not and merely references some older programs that did. I think it just uses a somewhat custom form of self-play reinforcement learning combined with MCTS.

[ Parent | on HN ]

Re:Alpha Go Zero: How and Why It Works by psb217 on Sunday November 05, 2017 @02:02PM (#15630952)

You are correct. There is no TD learning in AGZ. The value network is trained to directly predict the game outcome given the current state, and is not trained through "bootstrapping" based on the next state's value estimate.

[ Parent | on HN ]