Reducing Information Systems with Uncertain Real Value Attributes
Marzena Kryszkiewicz, Henryk Rybiński · Studies in fuzziness and soft computing · 2000
Information systems with real value multivalued attributes are considered. It is also assumed that some of the values describing objects in the system are uncertain. A procedure for quantization of sets of real numbers is presented. It transforms the information system into a reduced system, where the sets of reals are represented by ranges. The method generates ranges directly from the source table. As a result a set of optimal decision rules is provided.