Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!news.kei.com!sol.ctr.columbia.edu!howland.reston.ans.net!news.sprintlink.net!unify!sactoh0!stever
From: stever@sactoh0.sac.ca.us (Stephen A. Rickman)
Subject: ITRULE?
Message-ID: <Cy5zMF.Hq2@sactoh0.sac.ca.us>
Organization: Sacramento Public Access U*nix
Date: Mon, 24 Oct 1994 06:35:02 GMT
Lines: 14


Has anyone used an algorithm called ITRULE? According to a paper
I have, the algorithm develops a set of probability statements
from a database which can then be used to program the connections
and weights in a special non-learning NN. The claimed advantage
is that the behavior of the NN can be traced unambiguously to
particular rules. The method was developed by a couple of
researchers at JPL in the 80's. 

Any information would be appreciated.

-- 
Steve Rickman
stever@sactoh0.sac.ca.us
