Author: blass uri
Date: 08:33:39 04/13/98
Go up one level in this thread
On April 12, 1998 at 09:12:50, Pat King wrote: >I have been toying with the idea of using genetic programming to "grow" >variations.. populations of white move sequences vying against >populations of black sequences. Some manual analysis leads me to think >it may offer an advantage over AB for greater than 4 ply, and probably a >combination, using shallow AB to generate likely extensions to a given >variation, would work even better, although I have not done any work on >this yet. > I do not understand your idea of genetic programming I am sure the alpha-beta algotithm(I understand that you mean to this) is not a good algorithm for a long time because I see the computer program does not follow its main line and cannot see some combination that after a forced variation it sees in a short time. >Has anybody tried/seen anything like this? One stumbling block that I've >come across is that the losing population can prevent deep analysis by >favoring invalid moves that force evaluation at shallow depth. Can >anybody think of any others? Or is this old news, and if so, can >somebody point me to it? > >TIA > >Pat
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.