Classification Support Based on the Rough Sets Theory

Problems of knowledge analysis for decision systems by means of the rough sets theory are considered in this paper. Knowledge coming from experience concerns classification of data and is represented in a form of an information system. Application of the rough sets theory to the analysis of information systems enables a reduction of superfluous information and the derivation of a decision algorithm. These results are used to support a classification of new facts. An idea of using “the nearest” rules to support this classification is presented.