Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!ncar!gatech!psuvax1!psuvm!dbngmd13!dmswwu1c!zur072
From: ZUR072@DMSWWU1C.BITNET (Ulrich Kuehn)
Newsgroups: comp.ai.neural-nets
Subject: Re: Back Propagation problem..help needed.
Message-ID: <91059.131559ZUR072@DMSWWU1C.BITNET>
Date: 28 Feb 91 14:47:02 GMT
References: <1991Feb20.135431.27119@maths.tcd.ie>
Lines: 9
Organisation: Westfaelische Wilhelms-Universitaet, Muenster, Germany

In <1991Feb20.135431.27119@maths.tcd.ie> Thomas Murphy writes:
  Also, do I need thresholds, so far I have ignored them but are they
  important.

Yes, tese thresholds are very important for convergence of bp.  Yo can handle
them as a connection to a neuron which is always turned on, so all you have to
do is to use the learning algorithm on this weight for each neuron, too.

Ulrich Kuehn