关键词: 粗糙集, 三枝决策, 风险损失, 决策约简, 粒度分层

Abstract: The model of the theory of three-way decision-theoretic rough sets is used for the recognition of handwritten Chinese characters. A model of the theory of three-way decision-theoretic rough sets is set up which the regional classification is based on positive, negative and boundary. It can better reflect the recognition of handwritten Chinese characters classification on approximation. And define characteristic attribute of reduction of relative granularity entropy and the significance of attribute in offline handwritten Chinese characters recognition decision information system. If consider the recognition of handwritten Chinese characters from the point of view of the importance of attribute, it can design a kind of bottom up particle network structure based on logical operation between grain and grain, you can obtain more effective information. Simulation results show that, the method is feasible and effective.

Key words: rough set, three-way decision making, risk of loss, attribute reduction, vary granularity