【作者】 周创德； 田卫东； 周红鹃； 钱宏亮；

【Author】 ZHOU Chuang-de TIAN Wei-dong ZHOU Hong-juan QIAN Hong-liang School of Computer and Information, Hefei University of Technology, Hefei, 230009 Anhui, China

【机构】 合肥工业大学计算机与信息学院；

【摘要】 本文提出一种基于阈值约束的属性约简算法。算法先采用随机抽样的思想抽取部分数据以计算识别函数ΔA,然后根据A的特征设定阈值α,再用阈值α来指导对原数据集的属性约简以控制分辨矩阵元素的数量,从而获得较好的算法性能。实验结果表明,该算法的时空性能优于传统算法。Δ更多还原

【Abstract】 An attribute reduction algorithm based on threshold constraint, which makes use of randomly sampling methods, is proposed in this paper. The idea of this algorithm is to set the value of threshold α according to the features of identification function △ A computed from the subset dataset which is sampled from original dataset randomly, then to compute the reductions of original dataset using α in order to control the number of elements of discernable matrix and improve performance. The experiment shows that this algorithm outperforms the revisedly classic algorithm.更多还原