Rough set theory, proposed by Pawlak in1982, is an effective tool to deal with the vague, uncertain and incomplete data; Multi-granulation rough set is a new direction in rough set theory, which is based on multiple binary relations on the universe. The thesis aimed to study the knowledge reduction of multi-granulation rough set extension model in the interval valued information system.●The thesis defines the upper and lower approximation of multi-granulation in the interval valued information system under dominance relation. Besides, given multi-granulation rough set based on dominance relation, upper and lower approximation distribution reduction and its corresponding algorithm are listed in the thesis.●In the interval valued information system, the variable precision compatibility relation is also introduced. The thesis defines optimistic and pessimistic multi-granulation upper and lower approximation based on Variable precision compatibility relations. Further more, the thesis investigates some of its important properties and defines the upper and lower approximation distribution reduction. It also gives the algorithm of lower approximation distribution reduction.●In the interval valued information system, fuzzy rough set has been introduced. The thesis defines optimistic multi-granulation fuzzy rough set based on fuzzy compatibility relations. In addition, attribute reduction algorithm based on fuzzy positive region is put forward and its effectiveness got proved through the example given in the thesis. |