Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!news.duke.edu!concert!sas!mozart.unx.sas.com!saswss
From: saswss@hotellng.unx.sas.com (Warren Sarle)
Subject: Re: How to implement optimizing problems in NN?
Originator: saswss@hotellng.unx.sas.com
Sender: news@unx.sas.com (Noter of Newsworthy Events)
Message-ID: <Cy8sLw.13w@unx.sas.com>
Date: Tue, 25 Oct 1994 18:56:20 GMT
References:  <19941024113055.BKAMP@pi0192.kub.nl>
Nntp-Posting-Host: hotellng.unx.sas.com
Organization: SAS Institute Inc.
Lines: 33


In article <19941024113055.BKAMP@pi0192.kub.nl>, BKAMP@kub.nl  (Kamp B.) writes:
|> From the literature I recognize three main streams of NN-problem solving:
|> -classification
|> -prediction
|> -optimizing
|>
|> Regarding classification and prediction I have a rather clear picture, but
|> not regarding optimizing.

Classification and prediction are special cases of optimization.

|> A common problem in this area is the travelling salesman problem. I don't
|> understand how this type of problems can be implemented in a NN.
|> What kind/type of inputs and outputs does this require?
|> How is the optimum presented in terms of output?

See Cichocki & Unbehauen (1993), _Neural Networks for Optimization and
Signal Processing_, Wiley, who show how to compute almost anything
with neural nets. (Whether you _want_ to compute something with a
neural net is another matter.)

|> Can this kind of problem be solved by backprop or does one need
|> other algorithms?

One often needs other algorithms.


-- 

Warren S. Sarle       SAS Institute Inc.   The opinions expressed here
saswss@unx.sas.com    SAS Campus Drive     are mine and not necessarily
(919) 677-8000        Cary, NC 27513, USA  those of SAS Institute.
