You may want to look at using a classification tree. This will provide
"rules" that "predict" default rates in a loose sense,
rather than a model. However, in some cases the rules may be useful. For
example, in scoring a database of loan
applicants.
Tom
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
