Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!bt!uknet!gdt!aber!news
From: dbk@aber.ac.uk (Douglas B. Kell)
Subject: Re: "Population"?
Message-ID: <1994Sep3.153317.3779@aber.ac.uk>
Sender: news@aber.ac.uk (USENET news service)
Nntp-Posting-Host: pcfcej.dbs.aber.ac.uk
Organization: The University of Wales, Aberystwyth, U.K.
X-Newsreader: <WinQVT/Net v3.9>
Date: Sat, 3 Sep 1994 15:33:17 GMT
Lines: 18

In article <CvF7Jr.Euw@unx.sas.com> saswss@hotellng.unx.sas.com (Warren Sarle) writes:
>
>I recently encountered the word "ensemble" used in the statistical
>sense of "population", i.e. the set of all input vectors one wants
>to generalize to. Is this term widely used and recognized?
>

In statistical thermodynamics, based on work by Maxwell and Boltzmann, J. Willard Gibbs 
(of Gibbs Free Energy fame) developed the concept of an ensemble to mean a collection 
of particles possessed of the same energy (mean and time-averaged distribution). On this
basis it means a population but a population with specific properties (thermal and
thermodynamic equilibrium and thus a particular statistical distribution).

Best,
Douglas.
 


