Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!europa.chnt.gtegsc.com!gatech!news.sprintlink.net!redstone.interpath.net!sas!mozart.unx.sas.com!saswss
From: saswss@hotellng.unx.sas.com (Warren Sarle)
Subject: Re: Initial Weights with BP
Originator: saswss@hotellng.unx.sas.com
Sender: news@unx.sas.com (Noter of Newsworthy Events)
Message-ID: <DBEtsK.459@unx.sas.com>
Date: Sat, 8 Jul 1995 18:24:20 GMT
X-Nntp-Posting-Host: hotellng.unx.sas.com
References: <3tkmkv$imj@cnn.Princeton.EDU> <3tmb63$gfp@newsbf02.news.aol.com>
Organization: SAS Institute Inc.
Lines: 22


In article <3tmb63$gfp@newsbf02.news.aol.com>, jxidus@aol.com (JXidus) writes:
|> It isn't called a local minimum if it's the lowest point, it's the global
|> minimum. 

I don't know what the question was, but this answer is wrong. Any
global minimum is also by definition a local minimum. "Local" means
"within a sufficiently small neighborhood of the point in question".
Since a global minimum is minimal within _any_ neighborhood, it is
also minimal within any small neigborhood, so it is a local minimum.
I am not aware of any standard term for a local minimum that is
greater than the function value at some point outside the neighborhood
(a global minimum may not exist). I usually use "bad local mimimum",
which requires less typing than "suboptimal local minimum".



-- 

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.
