The attribute reduction of the information system based on new rough set

Attribute reduction is considered as an important preprocessing step for pattern recognition, machine learning, and data mining. The traditional rough set theory is mainly used to reduce the attributes and keep the lower approximation unchanged. In this paper we first give two forms of new rough sets: object-oriented rough set and attribute-oriented rough set, and then discuss their properties in detail. Based on the new models, this paper studies the attribute reduction of information system. At last it studies the attribute reduction of decision information systems by combining the old rough set and new rough set together.