A Connection Degree-Based Rough Sets Model and its Application to Knowledge Reduction

Knowledge reduction is one of the most important issues in rough sets theory. Many works are related to connection-degree-based rough sets model to treat incomplete information systems. However, the model is not used to knowledge reduction. In this paper, a connection-degree-based rough sets model is proposed and used to knowledge reduction. Two ways for knowledge reduction are proposed and the relationships between them are shown. A knowledge reduction algorithm is also shown for incomplete information systems. The algorithm is based on elementary discernibility matrix and the significance degree of an attribute. Examples are used to show that the validity of the algorithm.