I'm a bit confused about the difference between AMAF and RAVE (if there's one). As far as I understand, with AMAF, you sample in each playout (after it leaves the tree) the moves played (by both players), but only the first move at any position, then you reward all moves played either by a win or loss, depending on their colors.
I tried comparing this to RAVE, as described in many papers. There might be multiple definitions of RAVE, but the one that seems the most clear to me is this one (picture used because of math stuff): http://img352.imageshack.us/img352/443/bild1pb3.png Is it correct that with RAVE, after a playout (after the tree), only the siblings of the last node in the tree are updated accordingly? The main difference to AMAF would be that instead of a single array with values of AMAFsumReward and AMAFnumPlayed, each node keeps for each of its children separate variables that hold these values. When a playout is finished, only the values of these children are updated instead of the single array. I hope you're able to make any sense out of this, sometimes a text can be confusing when the writer is confused. ;p Cheers, ibd -- Sensationsangebot verlängert: GMX FreeDSL - Telefonanschluss + DSL für nur 16,37 Euro/mtl.!* http://dsl.gmx.de/?ac=OM.AD.PD003K1308T4569a _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
