Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!satisfied.elf.com!news.mathworks.com!news.kei.com!wang!news
From: emegged@f8maf1.jer.intel.com (Efi Megged)
Subject: Help on hidden layer
Organization: Intel Israel
Date: Wed, 15 Feb 1995 12:33:46 GMT
Message-ID: <3hssba$1unh@ilx018.iil.intel.com>
Sender: news@wang.com
Lines: 12

I want to use a neural net for  classification. The input is a picture
16*8 and the output is ~20 neurons my question, is there a way to
find the optimal (or even just a good estimate) size of the hidden
layer and if so is there a reference somewhere.
   Thanks in advance
 
----------------------
Not speaking for Intel 
Efi Megged
iNTEL ELECTRONICS ISRAEL
MailAddress EMEGGED@FAB8.INTEL.COM
----------------------
