Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ulowell!hawk!hbhatnag
From: hbhatnag@hawk.ulowell.edu (Himanshu Bhatnagar)
Newsgroups: comp.ai.neural-nets
Subject: Re: Training
Message-ID: <12506@swan.ulowell.edu>
Date: 30 Mar 89 04:38:45 GMT
References: <2698@sun.soe.clarkson.edu>
Sender: news@swan.ulowell.edu
Reply-To: hbhatnag@hawk.ulowell.edu (Himanshu Bhatnagar)
Organization: University of Lowell, CS Dept.
Lines: 15

In article <2698@sun.soe.clarkson.edu> spam@clutx.clarkson.edu writes:
>Has anyone come up with a good way to train a net
>without knowing a "target" in advance?  For example,
>if my net is in an undesireable state S1, and it is
>totally clueless about how to get to state S2, I would
				      ^^^^^^^^
I depends, if you know what the state S2 is or what it could be (heuristicslly)
then 
you could use the methods of Unsupervised Learning (Self Organsing Nets)
or Competitive learning... I could explain in detail a bit later, am in a 
hurry now...

>-Roger

Himanshu....