Rough set model based on Parameterized Probabilistic similarity relation in incomplete decision tables
Nguyen Do Van, Kôichi Yamada, Muneyuki Unehara · 2012
This paper discusses some extension of Rough set approach in Incomplete decision tables to deal with a problem of tolerance relation. Those approaches have been widely used to discover knowledge in incomplete information system. However, they also have their own limitation. In order to get more information from the relationship among objects, we propose a model called Parameterized Probabilistic Rough Set for incomplete decision tables. First we defined the probability of similarity between two objects if there is unavailable information. Then this probability is combined with a comparison based on available attribute values to derive a new relation.