Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!fs7.ece.cmu.edu!hudson.lm.com!godot.cc.duq.edu!news.duke.edu!news.mathworks.com!uhog.mit.edu!news.mtholyoke.edu!news.umass.edu!umassd.edu!umassd!s_vogt
From: s_vogt@umassd.cis.umassd.edu (Stefan Vogt)
Subject: Re: delta-bar-delta & delta-delta rule
Message-ID: <Cz30C8.94I@umassd.edu>
Sender: usenet@umassd.edu (USENET News System)
Organization: University of Massachusetts Dartmouth
X-Newsreader: TIN [version 1.2 PL2]
References: <39onfj$a2i@lorne.stir.ac.uk>
Date: Fri, 11 Nov 1994 02:31:18 GMT
Lines: 14


Peter Sincak (Guest) (psi@cs.stir.ac.uk) wrote:
: HI ,, 

: Does anybody have experience with Delta-bar-delta rule for BP?
: Can anybody explain me shortly how Delta-bar-delta was derived from 
: delta-delta ? Well reading Jacobs paper(Neural Networks Vol1, pp. 295-307
: 1988) I cant understand it. I would appreciate any commets. There is 
: part of it in Haykins book also. 

On page 301 there are the equation for updating the learning rate.
Where exactly is your problem?

Bye Stefan
