Uncertainty measure based on generalized rough sets covering

Uncertainty measure is a key issue of uncertain systems. In this paper, conditional entropy and increment information are introduced to discuss the two set relations in the same incomplete information systems based on covering approximation space. Two new kinds of measurements about the knowledge of rough set are presented. Furthermore the finer the set is, the lower the conditional entropy and increment information are, but the converse proposition does not hold. Finally, we discuss sufficient condition of the converse proposition.