Composing Approximations from Distribution Data under a Fuzzy Relation

Abstract

People often encounter some difficult situations to make their decisions due to information they can get less than that of they need. The theory of rough sets has been testified as an outstanding way to process incomplete and uncertain information. Fuzzy rough set generalized of classic rough set is to apply in non discretization information processing. Computing rough approximations is a necessary step for knowledge discovery by employing rough sets methodologies. Our taget of this research is to build the mechanism of composing approximations under a fuzzy relation instead of beginning from the initial point. For exploring the correlation of approximations computation, we respectively construct a decomposition and composition mechanisms for computing approximations of fuzzy rough concepts with Tcos-fuzzy equivalence relation. Under these mechanisms, some available located results were employed in approximations composition. A numerical illustration shows the composition of approximations based on a Tcos-fuzzy equivalence relation.