Rough Sets Consisting of Valued Objects in Data Tables with Missing Values

Abstract

We describe rough sets with values assigned to each element to deal with missing values in data tables under possible world semantics. It is clarified that with what value an object belongs to approximations. There exist possible tables that create elements of lower and upper bounds of indiscernibility relations, indiscernible classes, and approximations. This is the justification of this approach. The lower and the upper bounds of approximations are stricter than those of the previous work.