基于一种新的量化容差关系的变精度粗糙集模型

Variable Precision Rough Set Model Based on a New Quantitative Tolerance Relation

提出了一种新的量化容差关系,建立了变精度粗糙集模型并提出了基于重要度的属性约简算法。新的模型考虑了决策属性值对条件属性值概率分布的影响并且具有一定噪声数据处理和错误分类容许能力。最后通过实例表明了该模型具有处理含有空值数据问题的能力。

A new quantitative tolerance relation is designed, and then variable precision rough set model has been established and a reduction algorithm based on the significant degree is proposed. The new model takes into account the fact that the decision attribute value can affect the probability distribution of the conditional attribute value and has the capacity to deal with noise data and permit misclassification. Finally, an example shows that the model has the ability to deal with the problem which contains null data.