论文部分内容阅读
针对二进制可分辨矩阵属性约简方法在处理大数据集时的不足,首先给出两种二进制可分辨矩阵属性约简的定义,并证明这两个属性约简定义与正区域的属性约简定义是等价的;然后,给出对二进制可分辨矩阵按条件属性垂直划分后进行属性约简的方法;为了进一步降低空间开销,提出将垂直分解的二进制可分辨矩阵存于外部介质中,在约简过程中,仅将所需部分调入内存,由此设计启发式属性约简算法,其时间和空间复杂度的上界分别为O(∣C∣∣U∣2)和O(∣U∣2);最后,理论分析和实验结果验证了该算法的正确性和高效性.
Aiming at the shortcomings of binary discernable matrix attribute reduction method in dealing with large data sets, the definition of attribute reduction of two kinds of binary discernable matrices is given first, and the definition of attribute reduction of the two attributes is proved. Is equivalent; then, the method of attribute reduction is given after the binary discernable matrix is divided vertically according to the condition attributes; in order to further reduce the space overhead, it is proposed to store the vertically decomposed binary discernable matrix in an external medium, In the simple process, only the required part is transferred to the memory, so heuristic attribute reduction algorithm is designed. The upper bounds of time and space complexity are O (|C||U|2) and O (|U| 2). Finally, the theoretical analysis and experimental results verify the correctness and efficiency of the proposed algorithm.