分析了变精度粗糙集模型属性约简过程出现跳跃的原因,并给出消除跳跃现象的方法。探讨了基于分类质量、相对正域和决策类下近似的属性约简定义,并采用属性添加法对条件属性进行约简,约简过程反映了分类能力的变化。
The reason for bounce phenomena is analyzed based on attribute reduction in the variable precision rough set model, and the method to remove this phenomena is proposed in this paper. The proposition of attribute reduction based on the quality of classification, relative positive region and lower approximation is also discussed. Addition method is used to reduct attribute and the process reflects the change of the ability of classification.