Re: [S] sammon() function question

Prof Brian Ripley (ripley@stats.ox.ac.uk)
Tue, 24 Nov 1998 16:48:43 +0000 (GMT)


> Date: Tue, 24 Nov 1998 11:21:55 -0500
> To: s-news@wubios.wustl.edu
> From: Bill Shipley <bshipley@courrier.usherb.ca>
> Subject: [S] sammon() function question
>
> The Venables & Ripley MASS library includes a function called sammon().
> This function performs a nonlinear mapping of data from a higher dimensional
> space into a smaller (eg. 2) dimensional space. I give the reference for
> this at the end of the posting.
> My question concerns the relationship of this to non-metric multidimensional
> scaling. Both seem to do somewhat the same thing, but I don't know how
> (exactly) they differ. Is one superior to the other under certain conditions?
> Any leads about the sammon algorithm, or its comparison to other methods, is
> appreciated.

Ripley, B.D. (1996) Pattern Recognition and Neural Networks. CUP. Chapter 9.

Briefly, Sammon tries to preserve short distances, the usual non-metric MDS
tries to preserve the ordering of distances. Sammon is usually much faster.

-- 
Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

----------------------------------------------------------------------- This message was distributed by s-news@wubios.wustl.edu. To unsubscribe send e-mail to s-news-request@wubios.wustl.edu with the BODY of the message: unsubscribe s-news