Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!news.cac.psu.edu!news.pop.psu.edu!psuvax1!news.ecn.bgu.edu!feenix.metronet.com!net81.metronet.com!merriman
From: merriman@metronet.com (David K. Merriman)
Subject: Possibly dumb question: how to store state after training?
Sender: usenet@metronet.com (UseNet news admin)
Message-ID: <merriman.30.003FB2B7@metronet.com>
Date: Fri, 11 Nov 1994 14:01:49 GMT
Lines: 11
Nntp-Posting-Host: net81.metronet.com
Organization: some, but not much
X-Newsreader: Trumpet for Windows [Version 1.0 Rev B final beta #4]

I'm starting to get the hang of configuring NNs, but have run into a potential 
problem: how to store a net so that after it's trained, it can be re-loaded 
without having to go through training again.  Surely, there's a better way 
than going through and trying to save all the parameters for each and every 
connection; but what/how?

Dave Merriman

- - - - - - - - - - - - - - - - - - - - - - - - - - 
Finger merriman@metronet.com for PGP/RIPEM public keys and fingerprints. Unencrypted Email may be ignored without notice to sender.  PGP preferred.
Remember: It is not enough to _obey_ Big Brother; you must also learn to *love* Big Brother.
