论文部分内容阅读
在粗糙集理论中,分辨能力反映拥有知识的多少;为此,给出分辨能力相关概念、性质和计算方法,并提出基于相对分辨能力的约简定义,同时研究该约简定义与Hu差别矩阵约简之间的等价性,指出Hu差别矩阵约简可由相对分辨能力约简获得.为了进一步提高求解效率,通过减少约简过程中基数排序次数来提升效率,设计了相对分辨能力的约简算法,其时间复杂度为O(|C|~2|U|).实例分析和UcI中数据集的实验比较表明所提出的约简算法是有效的、可行的.
In the rough set theory, the ability of resolution reflects how much knowledge is owned. For this reason, the concepts, properties and calculation methods of resolution are given, and the definition of reduction based on relative resolving power is proposed. At the same time, the definition of reduction and Hu difference matrix The reducibility of Hu difference matrices can be obtained from the relative resolving power reduction.In order to further improve the efficiency of the solution, the efficiency of reducing the number of radix sorts in the reduction process is improved, and the reduction of relative resolving power Algorithm, the time complexity is O (| C | ~ 2 | U |) .An example analysis and experimental comparison of datasets in UcI show that the proposed reduction algorithm is effective and feasible.