Rough Set Theory (RST), proposed by researchers in 1982, is a mathematical tool for classification and analysis of imprecise, uncertain, incomplete or vague information and knowledge. Using various concepts in RST, classification of objects in various classes is done by removing irrelevant and redundant data by using attribute reduction algorithm. Attribute reduction using rough set theory is one of the key problems. Also, to find minimal attributes is a NP-hard problem. Attribute reduction is a preprocessing step in data mining, pattern recognition, machine learning, etc.