1.At last, we propose the concept of E-relative reduct of condition attributes and give its discernibility function computing approach.
最后给出条件属性的E-相对约简的概念及其区分函数求法。
2.and then comes up with a new attribute reduction algorithm with the help of discernibility matrix and logic operation.
通过分析多种属性约简算法，结合可辨识矩阵和逻辑运算，提出了一种属性约简算法；
3.The data processing time-consumed by discernibility of matrix is predicted, and the feasibility of data mining experiment is forecasted.
推算可预知分辨矩阵法处理数据所消耗的时间，对数据挖掘实验的可行性进行预测。
4.Discernibility ability index DI(a) of decision table is defined, some properties about DI(a) are presented.
根据决策表定义条件属性区分能力指数DI（a）的概念，给出DI（a）的若干性质。
5.To this issue, an attribute frequency reduction algorithm based on improved discernibility matrix is presented in this article.
针对这一问题，提出了一种基于改进可辨识矩阵的属性频率约简算法。
6.A novel rough set-based method followed by establishing a mix discernibility matrix is introduced for intrusion detection, and choose C4.
针对入侵检测问题，提出了构造混合辨别矩阵的方法，并用C4。
7.Aiming at these disadvantages, a new selected attribute standard based on magnitude of the discernibility element is proposed.
针对这些不足，提出基于差别元素的大小为新的属性选择标准。
8.In the algorithm, the function of attribute frequency is computed only using the discernibility matrix of data set.
该方法仅利用区分矩阵就可以计算出属性的出现频率函数值，计算简单。
9.At the second part of the thesis, an algorithm of attribute reduction based on binary discernibility matrix is presented.
其次提出了一种基于二进制差别矩阵的属性约简算法；
10.Heuristic knowledge reduction approach based on discernibility matrix and strong compressible set is proposed.
提出了基于区分矩阵与强等价集的启发式知识约简方法。