关键词: 单调性, 下近似保持, 下近似单调, 变精度粗糙集

Abstract: It is well-known that, the monotonicity plays an important role in attribute reduction of classical rough set. However, such property does not always hold in some generalization models. Variable precision rough set is a typical example. From this point of view, the definition of lower approximate monotonicity attribute reduction is presented in variable precision rough set model. In lower approximate monotonicity attribute reduction, the decision maker prefers to increase the lower approximate set rather than preserving the lower approximate set unchanged. The attribute reduction approach is also given to compute the reduct. The experiment results show that by comparing with lower approximate preservation reduct, the lower approximate monotonicity reduction not only increases the number of positive rules, decreases the number of boundary rules, but also increases the classification accuracy. It follows that, the lower approximate monotonicity reduction increases the certainties which are expressed by positive regions, and decreases the uncertainty coming from boundary region.

Key words: monotonicity, lower approximate preservation, lower approximate monotonicity, variable precision rough set