Author: Jay Scott
Date: 13:40:31 12/30/97
Go up one level in this thread
On December 29, 1997 at 21:23:03, Fernando Villegas wrote: >And then, what happened with Kon? Have you abandoned it or just >suspended? Don't you think to begin again that effort' Are you doing >that? In short, what a hell happened with Kon? And then I decided that I had learned all I wanted to in grad school and it would be a waste of time to put in another year of hard labor to get a piece of paper that said "PhD". So Kon is on the shelf, or rather, stored up in the attic. I learned a lot from the project, but the attic is where it belongs now. Some of the simplifying assumptions in the search algorithm design turned out to be major limitations. Eric Baum and Warren Smith designed an algorithm called BP which has more realistic assumptions, and it's amazingly complicated. See http://forum.swarthmore.edu/~jay/learn-game/methods/rational.html for my capsule summary and pointers. And my learning algorithm was somewhat weak too, witness the inability of Kon to learn to avoid all two-movers. I have a new, more sophisticated learning project on the burner now. And someday in the distant future I'll again have a suitable machine, and a new program will appear on the servers. Until then I'd prefer to shut up and keep working on it. :-) Jay
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.