home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.ai.neural-nets
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!sdd.hp.com!usc!cs.utexas.edu!qt.cs.utexas.edu!yale.edu!jvnc.net!newsserver.technet.sg!nuscc!nusunix3.nus.sg!isc40031
- From: isc40031@nusunix1.nus.sg (FOO JON HUI)
- Subject: Re: Meaning of deltas in hidden units
- Message-ID: <1992Dec18.152536.28947@nuscc.nus.sg>
- Sender: usenet@nuscc.nus.sg
- Organization: National University of Singapore
- X-Newsreader: TIN [version 1.1 PL6]
- References: <1992Dec15.162420.22854@Informatik.TU-Muenchen.DE>
- Date: Fri, 18 Dec 1992 15:25:36 GMT
- Lines: 15
-
- Boris Baginski (baginski@Informatik.TU-Muenchen.DE) wrote:
- :
- : But the question is: what do those values express for the hidden units and
- : for the input units (can be calculated easily, are not used) ?
- : We explored that these deltas take on values up to 50 (+/-) or so, but we only
- : use inputs and targets and activation states in the range of 0-1.
-
- Unless I've been doing things wrong, the inputs don't really have to be
- in th 0-1 range, since at the input they are multiplied by a weight of
- real value. Of course, if you use the sigmoid function for backprop,
- your outputs must be between 0-1.
-
- Jon.
- National University of Singapore.
-
-