Author: Ulrich Tuerke
Date: 07:51:56 05/08/98
Go up one level in this thread
Hi Don, thanks for your good and clear presentation. I too think that pre-processing is still done in most of the current chess programs, - more or less. I am considering the possibility of a moderated pre-processing, i. e. apply the time-consuming pre-processing evaluation parts not at the root but to nodes at depth - say - the half of the aimed search depth. Thus, one would still have the main advantage, saving a lot of calculation time, because the number of pre-processed nodes would still be a microscopic fraction of the leaf nodes. But on the other hand, the preprocessed knowledge would be applied to nodes being far closer to the leaf nodes (in your examples "only" 30 ply instead of 60). Opinions ?
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.