Dual Rough Approximations in Information Tables with Missing Values

A method of possible equivalence classes has been developed under information tables with missing values. To deal with imprecision of rough approximations that comes from missing values, the concepts of certainty and possibility are used. When an information table contains missing values, two rough approximations, certain and possible ones, are obtained. The actual rough approximation lies between the certain and possible rough approximations. The method gives the same results as a method of possible worlds. This justifies the method of possible equivalence classes. Furthermore, the method is free from the restriction that missing values may occur to only some specified attributes. Hence, we can use the method of possible equivalence classes to obtain rough approximations between arbitrary sets of attributes having missing values.