Author: Marc-Philippe HUGET
Date: 07:01:47 03/09/99
Go up one level in this thread
On March 09, 1999 at 09:55:11, Matt Frank wrote: >On March 09, 1999 at 09:32:24, Marc-Philippe HUGET wrote: > >>Hi, >> >>I know book learning is important in order to improve opening quality but I am >>wondering I just have to store position and weighted score. I think it is not a >>good idea. Let us suppose my program plays 10 games against, say, Fritz 5, it >>looses 10-0, so for each position I decrease score, and now my program plays 10 >>games against super dummy 1.0 and I wins 10-0. So I increase score. What is the >>conclusion ? If my program plays 6 games with e4 and wins 3 games, I don't >>change the score for this position however I am not sure this move is so good. I >>say e4 but it is the same thing for deeper moves. >> >>What are exactly the advantages of book learning and how exactly code the book >>learning in order to improve openings. I think position and score are not >>enough. Ideas ? >> >>mph > >I'd say that depth of search, evaluation, and if the evaluation proceded to >fininsh; final result are important in the learning function. I would suggest >that the learning should be focused on high level games and if necessary the >analyses used for learning should be constrained to only high level opponents. >Hope this helps. > >Regards, >Matt Frank Yes, but it is not very easy to know if your opponent is good or not. What do you think of book learning where we have position, score and best move for this position. While this best move is not refuted, we store this move, so the program only plays this move for this position. mph
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.