论文部分内容阅读
针对混合值不完备系统,提出一种基于双邻域粗糙集模型的分类方法。首先定义一个新的不确定距离度量函数——联系度距离函数,进而建立基于联系度距离函数的双邻域粗糙集模型;其次,基于所建立的模型,讨论了该模型的属性约简算法,并给出了基于属性约简、覆盖约简的双邻域粗糙集规则学习分类算法;最后,通过多个UCI数据集进行了实证分析,结果表明本文提出的分类算法是客观有效的,特别是在缺失值较多的情况下其优势更加明显。
Aiming at the incomplete mixed value system, a classification method based on two-neighborhood rough set model is proposed. Firstly, a new measure function of distance between uncertainties and contact degree is defined, and then a two-neighborhood rough set model based on the distance function of contact degree is established. Secondly, the attribute reduction algorithm of the model is discussed based on the established model. Finally, the empirical analysis based on multiple UCI datasets shows that the classification algorithm proposed in this paper is objective and effective, especially In the case of more missing value its advantage is more obvious.