I'd really like a good tutorial for combining MCTS with NNs, but this tutorial didn't help me at all really. MCTS is easy enough to learn on its own, but the details of connecting it to NN learning is still mysterious to me after having read this.
Of course I could go and study the LeelaZero or minigo code, but I was hoping for a nice readable introduction.
I think there is a chapter dedicated to this in the newest draft of Rich Sutton's Reinforcement Learning book, available at his personal page. Haven't read that chapter though.