I will appreciate very much receiving your thoughts on the
Build a fuzzy model (say a TSK model) using 20 variables, among
which 10 are continuous numerical variables (e.g., balance,
purchase amount etc) and 10 are categorical variables (e.g.,
occupation, zip code, etc).
Where is the difficulty?
The difficulty is in the incorporation of the categorical
variables into the model. There are various approaches to use
categorical variables in traditional regression models such as
linear and logistic, and these approaches can also be used
But how to use them in fuzzy models? A possible solution may be
to construct segmentations (like tree) using the categorical variables
and then build models in each segmentation. But you will see that this
approach is not practical when the number of categorical variables is
large (say 10) and the numner of the items is large (such as zip code).
Sr. Staff Scientist
HNC Software Inc.
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 email@example.com
(2) To unsubscribe from this mailing list, send a message body of
"UNSUB FUZZY-MAIL" or "UNSUB FUZZY-MAIL firstname.lastname@example.org"
(3) To reach the human who maintains the list, send mail to
(4) WWW access and other information on Fuzzy Sets and Logic see
(5) WWW archive: http://www.dbai.tuwien.ac.at/marchives/fuzzy-mail/index.html
This archive was generated by hypermail 2b30 : Fri Jul 06 2001 - 22:21:22 MET DST