Hi.
And one for chess players.
You have window of posible outcomes.
And correlate winrate with points. (Random komi)
What MCTS really does is:
for each node find posible moves
try some random play-outs to get winrates
prune bad, expand good
go deeper.
propagate winrate upstream (?min/max, confidence)
Now correlate winrates to points
for each node find posible moves
try some random play-outs to get winrates=points
prune bad, expand good
try to go deeper.
propagate winrate=points upstream (?min/max, confidence)
Does this sound familiar?
Leon.
_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
