粗糙集中离散化算法的研究

Research on Discretization Algorithms in Rough Sets

对目前粗糙集的离散化算法进行了分类讨论,重点分析了基于信息熵的离散化算法的理论基础以及实现步骤,并就该算法对于同一属性在不同样本数据集上的应用情况进行了分析.实验表明,该算法对于部分属性具有数据敏感性,当选择这些属性作为依据时会影响系统的决策能力.

Based on classification and discussion of current discretization algorithms,the theoretical basis and accomplishment steps of discretization algorithm using information entropy were analyzed in details,and applications of the algorithm in different sample sets of the same attribute were presented.The experimental results show that the algorithm has sensitivity to some attributes and can affect the decision ability of the system when chosen these attributes.