CLASS-L Archives

December 2001


Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Art Kendall <[log in to unmask]>
Reply To:
Classification, clustering, and phylogeny estimation
Sat, 15 Dec 2001 07:31:24 -0500
text/plain (1519 bytes) , Arthur.Kendall.vcf (140 bytes)
For the 76 election, I did my dissertstion using a 205 * (10*10) SINDSCAL.  I
just retired and everything is in boxes, but over time I intend to look at all
my VAX tapes to see what should be preserved.   At that time I translated
SINDSCAL to VAX Fortran which allows you to use PARAMETER statements and
compile to fit your problem.

Is this the kind of thing you are looking at? Are you more interested in
single matrix MDS?

I have worked with large data sets from Social Security, IRS, DoD, Medicare,
Census, etc.  I may very well come across needs in my future consulting.
(e.g., I was approached by the National Statistics Office of the UK.)

In the 70's I did a lot of clustering of counties, I have tossed around the
idea of doing 1) clustering of all the counties and/or 2) generating R
matrices for each county and INDSCALing those data.

David Dubin wrote:

> Mark Rorvig (Univ. North Texas), Michael Trosset (College of William and
> Mary), and I have begun a project on large-scale applications of
> multidimensional scaling. We are interested in hearing from other
> researchers working with large amounts of data, who might benefit from the
> use of the system we propose to build.
> Mark will post a URL to an overview of our project soon. In the mean time,
> please direct inquiries and statements of interest to my attention
> ([log in to unmask]) and to Mark Rorvig ([log in to unmask]).
> Good holidays to you all.
> Dave Dubin
> Senior Research Scientist
> Information Systems Research Lab
> University of Illinois at Urbana-Champaign