Re: basic BayesCraft training?

From: Peter de Blanc (peter.deblanc@verizon.net)
Date: Thu Mar 17 2005 - 18:57:49 MST


Something that might be useful as a learning tool would be a game (like
Zendo, maybe) that uses Bayesian Scoring. Or it could be simpler, like
guessing the next in a sequence in binary, or red-green-blue. You could
generate good sequences by creating random expressions in an internal
turing-complete interpreted language and evaluating them from n=1 to 50,
say.



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:50 MDT