Author: José Carlos
Date: 00:07:59 05/06/03
Go up one level in this thread
On May 05, 2003 at 19:47:22, Rex wrote: >On May 05, 2003 at 14:32:36, Robert Hyatt wrote: > >>On May 05, 2003 at 13:07:06, Steven Walton wrote: >> >>>which is better to do - because i think if you continue to import games from >>>somewhere you will run out fo hard disk space but learning , it not taking up >>>much space on the disk -- tell me if im wrong . >>> >>> >>>steve. >> >> >>Importing is _way_ faster. Otherwise you have to play each opening and >>discover for yourself whether it is good or bad. > > >I dont agree. One engine my loose while playing a particular opening when >another engine will win each and everytime with that same opening. Each engine >is different to some % with the opening and learning created by that engine is >just as important. I do agree some of the weaker openings are good to import >the learning. Hey save a game of loosing then. Learning is a slow process. Starting from an empty book, I might take years to create smething comparable to professional hand-tuned books, just by automatic learning. At least, with the "known" technics, or to be precise, the ones I know or can figure. If you think about it, for every lost game, the program should try to figure where was the mistake, see if it can find a better move, try it and evaluate the whole tree again. If the losing mistake was in the endgame then thing gets really difficult... José C.
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.