> The tiling is simply a tracking of how effect a move is when combined > with a specific follow-up move. Near the start of a simulation, this > would match RAVE values. Deep in a simulation, it's highly situational > and based on which follow-up moves remain open. > > I hope that helps!
Almost :-) It is still a bit fuzzy, but I think I'll wait for another paper on the subject before trying to apply any more brain power to it. Just skimming it again, their 57% win rate does not seem to allow for the extra CPU work required for doing the tiling. That is reasonable - you don't want people to waste time optimizing before they publish algorithms - but it does suggest it may turn out to be no better use of your CPU cycles than simply doing more dumb playouts. Darren >>> A just published paper about learning MC policies: >>> http://hal.inria.fr/inria-00456422/fr/ >>> It works quite well for Havannah (not tested on hex I think). >> >> I struggled with this paper ("Multiple Overlapping Tiles for Contextual >> Monte Carlo Tree Search"), as it wasn't clear to me what a "tile" was. >> Specifically I couldn't work out if they were 2d patterns of >> black/white/empty, or are they are a sequence of moves (e.g. joseki, >> forcing moves, endgame sente/gote sequences, etc. in go)? Or perhaps >> something else altogether? >> >> While I wear the dunce's cap and stand in the corner, is some kind soul >> able to explain the idea in go terms? -- Darren Cook, Software Researcher/Developer http://dcook.org/gobet/ (Shodan Go Bet - who will win?) http://dcook.org/work/ (About me and my work) http://dcook.org/blogs.html (My blogs and articles) _______________________________________________ Computer-go mailing list [email protected] http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
