基于简化差别矩阵的增量式属性约简

新对象添加到决策表后,已有的属性约简将会发生改变,需要对其动态更新.为此,首先给出简化决策表和简化差别矩阵的定义,并证明了基于简化差别矩阵的属性约简与正区域的属性约简是等价的；然后,分析增量对象的不同情况,将增量属性约简映射到简化决策表上来实现,由此设计基于简化差别矩阵的增量式属性约简算法,并对算法进行改进；最后,利用实例和实验验证了所提出算法的正确性和高效性.

Incremental Attribute Reduction Based on Simplified Discernibility Matrix

It is necessary that the attribute reduction of decision table will be updated dynamically, when the decision table object is updated and the original attribute reduction may be changed. Firstly, the definitions of the simplified decision table and attribute reduction based on simplified discernibility matrix are proposed. It is proved that attribute reduction acquired from the definition is equivalence to attribute reduction based on positive region. Secondly, the different cases of added object are analyzed, and incremental attribute reduction is mapped to the simplify decision table to achieve. Based on above, the attribute reduction algorithm based on simplified discernibility matrix is designed and the algorithm is further improved. Finally, both of example and experiment results show that the algorithms are effective.