Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!know!zaphod.mps.ohio-state.edu!uakari.primate.wisc.edu!aplcen!haven!mimsy!brillig.cs.umd.edu!grange
From: grange@brillig.cs.umd.edu (Granger Sutton)
Newsgroups: comp.ai.neural-nets
Subject: Re: Input scaling alters my results; why?
Summary: Normalized input vector lengths
Message-ID: <26019@mimsy.umd.edu>
Date: 14 Aug 90 19:29:54 GMT
References: <473@array.UUCP> <3200@gmdzi.UUCP>
Sender: news@mimsy.umd.edu
Reply-To: grange@brillig.cs.umd.edu (Granger Sutton)
Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742
Lines: 10

I looked at this phenomenon briefly a while ago. In addition to what other
people have already said the lengths of the input vectors (the activations
of the input nodes when viewed as a vector) seemed to have an impact on
learning. Specifically, if all the input vectors are the same length learning
times seem to be faster and more uniform (smaller variance). For binary
input vectors -1,1 coding gives you input vectors that are all the same
length.

Granger Sutton
grange@brillig.cs.umd.edu