home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.ai.neural-nets
- Path: sparky!uunet!europa.asd.contel.com!howland.reston.ans.net!wupost!spool.mu.edu!agate!doc.ic.ac.uk!clss1.bangor.ac.uk!sees.bangor.ac.uk!paulw
- From: paulw@sees.bangor.ac.uk (Mr P Williams (AD))
- Subject: Re: Questions about sigmoids etc.
- Message-ID: <1992Dec11.093325@sees.bangor.ac.uk>
- Keywords: Sigmoids, output layers
- Sender: paulw@sees.bangor.ac.uk (Mr P Williams (AD))
- Organization: University of Wales (Bangor)
- References: <waugh.723705045@probitas> <1992Dec8.161935@sees.bangor.ac.uk> <1g4abhINN7h4@corax.udac.uu.se>
- Date: Fri, 11 Dec 1992 09:33:25 GMT
- Lines: 33
-
- In article <1g4abhINN7h4@corax.udac.uu.se>, crwth@Merlin.DoCS.UU.SE (Olle Gallmo) writes:
- > In article <1992Dec8.161935@sees.bangor.ac.uk>, paulw@sees.bangor.ac.uk (Mr P Williams (AD)) writes:
- > >
- > >
- > > Hello,
- > > For backpropagation networks (i.e. Rumelhart ,Mclelland and Williams), it
- > > is neccessary to have a monotonically increasing, DIFFERENTIABLE function as
- > > the output
- >
- > Differentiable, yes, but to my knowledge there is nothing in the algorithm that
- > requires the transfer function to be monotonically increasing.
- >
- > >
- > > Any constructive criticism welcomed by me on this; I'm always willing to learn!!
- > >
- > >
- > > Paul
- >
- > Ditto!
- >
- > /Crwth
- >
- > ---- "If God is perfect -- why did He create discontinuous functions?" ----
- > Olle Gallmo, Dept. of Computer Systems, Uppsala University
- > Snail Mail: Box 325, S-751 05 Uppsala, Sweden
- > Email: crwth@DoCS.UU.SE
-
-
- Hi,
- I thought that the algorithm required unique values for x in f(x)?
-
- Paul
-
-