An important success factor for KDD lies in the development and
integration of methods for supporting the construction and execution
of KDD processes. There are three crucial aspects in this context:
the (incremental) development of a precise problem description; the
decomposition of this top level problem description into manageable
and compatible (reusable) subtasks; and the selection and combination
of adequate algorithms -- based on data characteristics among others
-- for solving these subtasks. In this talk, I outline the basic
principles and methods of the User Guidance Module that provides
support for both top-down problem refinement and decomposition as well
as bottom-up data analysis and algorithm selection. Bottom-up data
analysis is based on a collection of statistical and information
theoretic measures.
Date: Wed., May 20; Time: 4:15-5:30PM; Place: Cordura 100
Return to seminar schedule.