Many Faces of Go is MC + expert system (plus local search, etc). The reason I won the world championship in 2008 is because I implemented MCTS but incorporated the old Many Faces expert system move generator and ranking. This is pretty slow (a few hundred positions a second), so when the tree part of MCTS got a node up to about 100 visits, the Many Faces move generator was called and it applied a bias to the moves to favor moves that look good to the expert system. This was stronger than all the other pure MCTS programs.
Now the other MCTS programs are stronger because they incorporate more go knowledge, either through machine learning from expert games or from strong player input. I don’t think MCTS is stagnating. I think DH Brown is correct about how very much more difficult it is to climb the higher ranks. The rate of progress is about the same, but the rate of rank improvement is much, much slower. Many Faces has been stagnating because I have hardly touched the engine in the last 18 months. That’s changing soon. David From: Computer-go [mailto:computer-go-boun...@computer-go.org] On Behalf Of Stefan Kaitschick Sent: Friday, September 04, 2015 8:55 AM To: computer-go@computer-go.org Subject: Re: [Computer-go] re comments on Life and Death So far I have not criticised but asked questions. I am a great fan of the expert system approach because a) I have studied go knowledge a lot and see, in principle, light at the end of the tunnel, b) I think that "MC + expert system" or "only expert system" can be better than MC if the expert system is well designed, c) an expert system can, in principle, provide more meaningful insight for us human duffers than an MC because the expert system can express itself in terms of reasoning. (Disclaimer: There is a good chance that I will criticise anybody presenting his definitions for use in an expert system. But who does not dare to be criticised does not learn!) MC is currently stagnating, so looking at new (or old discarded) approaches has become more attractive again. But I don't think that a "classic" rules based system will be of much use from here. It is just too far removed from MC concepts to be productively integrated into an MC system. And no matter what, MC has to be the starting point, because it is so much more effective than anything else that has been tried.What you are left to work with, is the trail of statistics that MC leaves behind. That is the only tunnel with a possible end to it that I see. And who knows, maybe someone will find statistical properties that can be usefully mapped back to human concepts of go.
_______________________________________________ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go
