Re: Fuzzy C-Means convergence

Earl Cox (ecox@metus.com)
Tue, 12 Oct 1999 04:04:55 +0200 (MET DST)

Will,

It means that if you select the same number of initial clusters and seed
the cluster centers with the same values, you will get the same results from
run to run against the same data set. Fuzzy Clustering with FCM is
deterministic. If you select a different number of clusters or randomly seed
the cluster centers then you will, most likely, get different clusters for
each different set of FCM parameters. Bezdek showed that the FCM will always
converge -- that is, you get a stable response in any single run.

Earl

Will Dwinnell wrote in message <37FF74D8.57B6@compuserve.com>...
>I have read that fuzzy c-means has been proven to converge for all c,
>from c=1 to c=infinity. Does this mean that fuzzy c-means will always
>product the same result (from run to run on the same data), or does it
>mean that fuzzy c-means will (in any single run) settle down to a stable
>response?
>
>Will Dwinnell
>predictor@compuserve.com

############################################################################
This message was posted through the fuzzy mailing list.
(1) To subscribe to this mailing list, send a message body of
"SUB FUZZY-MAIL myFirstName mySurname" to listproc@dbai.tuwien.ac.at
(2) To unsubscribe from this mailing list, send a message body of
"UNSUB FUZZY-MAIL" or "UNSUB FUZZY-MAIL yoursubscription@email.address.com"
to listproc@dbai.tuwien.ac.at
(3) To reach the human who maintains the list, send mail to
fuzzy-owner@dbai.tuwien.ac.at
(4) WWW access and other information on Fuzzy Sets and Logic see
http://www.dbai.tuwien.ac.at/ftp/mlowner/fuzzy-mail.info
(5) WWW archive: http://www.dbai.tuwien.ac.at/marchives/fuzzy-mail/index.html