Author: Peter Fendrich
Date: 06:52:05 06/29/04
Go up one level in this thread
On June 28, 2004 at 14:49:42, Dan Andersson wrote: >>I'm doing similar things with GA right now. The difference between GA and >>Simulated Annealing is probably of not very important here. >> > There are considerable and fundamental differences between Genetic Algorithms >and Simulated Annealing. GA is searching the whole of the search space and is >capable of finding non-obvious approximate optimums. While SA is a heuristic >Monte-Carlo based optimization. > Which works best in practice for Chess Autotuning could depend on how close to >a 'real' understanding of practical chess evaluation we have in the chess >program at the moment of the test. > A mainly linear formula should be easy for both GA and SA. While a discontinous >partially ordered set would probably favour GAs. > >MvH Dan Andersson I'm well aware of the differences but in this context I think the quality of the result is less about GA versus SA than other issues like the quality of the the test setup. /Peter
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.