| Re: D vector pattern classification tips [message #5012] |
Wed, 13 September 1995 00:00 |
velt
Messages: 19
Registered: June 1994
|
Junior Member |
|
|
In article <Pine.SOL.3.91.950912152226.18036A-100000@chroma.mbt.washington.edu>, Russ Welti <rwelti@chroma.mbt.washington.edu> writes:
> I am doing Web and bibliography searches on clustering
> algorithms, with limited success.
> Has anyone done any of this before or have a good
> pointer they could offer? A favorite textbook? A web page?
Textbooks:
J. Hartigan: Clustering Algorithms. New York: John Wiley & Sons, 1975
A Jain and R Dubes. Algorithms that cluster data. Englewood Cliffs,
NJ: Prentice Hall, 1988.
Clutering algorithms in the IDL user library:
- Join (Hierarchical clustering);
- Kmeans (partitioning). The routine in the user's library is not
a standard k-means, but something that uses the same functional
as k-means does.
It appears that k-means does not exist in IDL 4.0, but I haven't
verified that.
I have written an IDL version of fuzzy c-means, which is the
fuzzy version of k-means. I can post it if anyone desires.
If you don't know how many clusters you have, there are ways of
using standard clustering techniques to detect the number of
classes, such as minimizing the Davis-Boulden index.
Robert Velthuizen
Digital Medical Imaging Program,
H. Lee Moffitt Cancer center and Research Institute;
Dept. of Radiology, University of South Florida.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
