Author: Dann Corbit
Date: 20:35:44 01/13/00
Go up one level in this thread
On January 13, 2000 at 23:15:02, Robert Hyatt wrote: [snip] >But allowing him to play every "personality" in a program that has roughly >8,000 (Hsu's number) tunable evaluation parameters would take a _long_ time. >(Hint: evaluate 8000! [factorial]) How many games per personality would be >required? (then think about that number * 8000!). It might be possible to do a non-linear optimization. If the functions are linear, even simplex might be used. It would certainly take a lot of games to make decisions that are accurate, though.
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.