Re: fundamental problems for reinforcement

[Weimin Xiao <ranxiao@xxxxxxxxxxxx>]

From: Ives Steglich <dalini@xxxxxxxxxxxxxxxxx>
...
a) i need a way to detect the end of a game
b) to rate it in that way that i can say which play has won

:- I have Wilcox's EzGo (6.1995 version) and found even though the program
will play to very end of a game, the program finds game score and game end
fairly well. HandTalk even will give up at mid-game. If your reinforcement
learning can be performed off-line, then only thing remained is to let the
commercial programs (or some other programs that will give out good end-game
evaluation) read existing game records, then write out final results.

Weimin