Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ukma!rutgers!att!cbnewsj!veenu
From: veenu@cbnewsj.ATT.COM (veenu.r.rashid)
Newsgroups: comp.ai.neural-nets
Subject: Re: Minimum # of internal nodes to form boolean function
Message-ID: <1383@cbnewsj.ATT.COM>
Date: 12 Oct 89 18:45:51 GMT
References: <1989Oct12.050402.9790@boingo.med.jhu.edu>
Reply-To: veenu@cbnewsj.ATT.COM (veenu.r.rashid,mt,)
Distribution: usa
Organization: AT&T Bell Laboratories
Lines: 24

In article <1989Oct12.050402.9790@boingo.med.jhu.edu> heath@cs.jhu.edu (David Heath) writes:
>
>  I've been told that Robert Heicht Neilson recently proved that any
>boolean function of n inputs to n outputs can be realized with a neural
>net having n input nodes, n output nodes and 2n-1 intermediate nodes
>(a total of 3 layers). Is there any truth to this statement? Please
>forgive me if this has been discussed here before.
>

Yes, this is true as a consequence of simple combinatorics.  Does anyone know the
number of intermediate nodes (hidden layer) necessary to realize a real valued function
[0..1] of n inputs to m outputs.  Is there a functional description of such a net
anywhere?


Inquiring minds would like to know...


>
>--------------------------------
>Dave Heath	heath@cs.jhu.edu


Navruze Rashid      ruze@mtfmi.att.com    -or-    veenu@cbnewsj.att.com