优势关系下广义决策约简和上近似约简

Generalized Decision Reduction and Upper Approximation Reduction Based on Dominance Relation

论文定义了决策表的优势关系下广义决策约简和上近似约简,给出了优势关系下广义决策约简和上近似约简的判定定理和辨识矩阵。同计算优势关系下上近似约简的辨识矩阵相比,计算优势关系下广义决策约简的辨识矩阵的时间复杂度低,由于论文已证明优势关系下广义决策约简和上近似约简是等价的,因此,可以利用优势关系下广义决策约简的辨识矩阵计算优势关系下广义决策约简和上近似约简。

Generalized decision reduction and upper approximation reduction based on dominance relation have been defined.It is proved that a generalized decison reduction based on dominance relation is equivalence to upper approximation reduction based on dominance relation.The judgement theorems and discernibility matrixes with respect to generalized decision reduction and upper approximation reduction based on dominance relation are established,from which we can obtain algorithms for finding generalized decision reduction and upper approximation reduction based on dominance relation.Compared with the algorithm for finding a discernibility matrix with respect to upper approximation reductlon,the time complexity of the algorithm for finding a discernibility matrix with respect to generalization decision reduction is lower.So the discernibility matrix with respect to generalization decision reduction can be used to find upper approximation reducts and generalized decision reducts.