What is your criterion for good features, then? Features that maximize your clustering goodness criterion (clusters such that items in clusters are closer to each other than to items in other clusters, or something of the sort)?
How about Friedman & Meulman, if you allow different feature subsets for different clusters: http://www.rss.org.uk/publications/docs/050504.pdf
Regards,
- Kari Torkkola
shannon wrote:
To clarify I am interested in feature selection in the unsupervised or clustering problem.
I have papers by DeSoete; Fowlkes, Gnanadesikan, Kettenring; DeSarbo, Carroll, Green. Any other suggestions?
Thanks again. Bill ---
Joint Meeting of the Interface and Classification Society of North America
http://ilya.wustl.edu/~kleint/if_csna_2005
William D. Shannon, Ph.D.
Associate Professor of Biostatistics in Medicine Division of General Medical Sciences and Biostatistics
Washington University School of Medicine Campus Box 8005, 660 S. Euclid St. Louis, MO 63110
Phone: 314-454-8356 Fax: 314-454-5113 e-mail: [EMAIL PROTECTED] web page: http://ilya.wustl.edu/~shannon
