利用差别矩阵构造决策树  

Construction of decision tree with discernibility matrix

作　　者：高静[1] 韩智东[2] 

机构地区：[1]首都经济贸易大学信息学院,北京100070 [2]中国银联北京信息中心,北京100193

出　　处：《计算机工程与应用》2011年第33期18-21,共4页Computer Engineering and Applications

基　　金：国家社科项目(No.11CYY020);教育部项目(No.10YJC740069);首都经济贸易大学项目(No.00791056721630;No.00791154210150)

摘　　要：分析了基于正区域、基于粗糙边界和基于依赖度的属性选择标准的关系,证明了这三种属性选择标准彼此等价。以正区域的属性选择标准为代表,分析了基于正区域的决策树生成算法的优点和不足。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。用新的属性选择标准生成的决策树一般具有叶子数目较少,叶子的平均深度也较小,且叶子具有较强的泛化能力。用一实例说明了新的属性选择标准的优越性。The relationship between selected attribute standards based on positive region,based on rough bound and based on attribute dependency is analyzed.It is proved that the three kinds of selected attribute standards are equivalent to each other.Advantages and disadvantages of algorithm for constructing decision tree based on positive region are analyzed.Aiming at these disadvantages,a new selected attribute standard based on magnitude of the discernibility element is proposed.The decision tree constructed with the new standard of attribute selection has the following characteristics：fewer leaf nodes,fewer levels of average depth,better generalization of leaf nodes.An example is used to illustrate the advantages of this new selected attribute standard.

关 键 词：决策树 粗糙集 正区域 粗糙边界 依赖度 差别矩阵 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]