| | |
Summary: Weitere Kolloquiumstermine sind im WWW abrufbar.
Fakultät II Informatik, Wirtschafts- und
Rechtswissenschaften
Department für Informatik
Informatik Kolloquium
Am Donnerstag, dem 16. Juni 2011, um 16:15 Uhr hält
Dr. Michal Grabowski
Warsaw School of Computer Science
einen Vortrag mit dem Titel
Metrics and similarities in modeling dependencies between
continuous and nominal data
Der Vortrag findet im OFFIS, Escherweg 2, Konferenzraum F02 statt.
Abstract:
In its analytical paradigm, classification theory investigates continuous data only. When we are faced with
a mixture of continuous and nominal attributes in data records, difficulties are emerging. Usually, the
analytical paradigm is maintained by treating nominal attributes as continuous ones via numerical coding
of nominal values (often a bit ad hoc). We propose a way of keeping nominal values within the analytical
paradigm with no pretending that nominal values are continuous.
The core idea is that the information hidden in nominal values influences the metrics (or the similarity
function) on records of continuous and nominal data. An adaptation via genetic algorithms finds the rele-
|
