News for hackers. Stuff that matters.
Re:An Introduction to Counterfactual Regret Minimization by eutectic on Friday February 03, 2017 @12:15PM (#13561351)
Is there a version of CFR for differentiable learners like neural networks?
Re:An Introduction to Counterfactual Regret Minimization by gwern on Friday February 03, 2017 @04:12PM (#13563362)
The averaging part makes it sound like the usual RL self-play against regular checkpoints of oneself.