基于属性重要度的ID3改进算法  

Improved ID3 algorithm based on attribute importance

作　　者：邹永贵[1] 范程华[1] 

机构地区：[1]重庆邮电大学计算机科学与技术学院,重庆400065

出　　处：《计算机应用》2008年第B06期144-145,149,共3页journal of Computer Applications

摘　　要：ID3算法是数据挖掘中最经典的分类算法。该算法偏向于选择取值较多的属性,而属性值较多的属性不总是重要的,从而影响了分类预测的高效性。通过对ID3算法的研究,依据属性重要度粗糙集理论的思想,对经典的ID3算法做了相应的改进,改进后的ID3算法(AIID3),提高了算法的决策效率。最后的实例及应用表明,改进的算法更有效,更快速。ID3 algorithm is a classical algorithm in data mining. This algorithm inclines to the attribute with more values, which affects the efficiency of classification and prediction. Via the research on ID3 algorithm, according the thought of attribute importance degree of rough set, the algorithm was improved. The decision efficiency of the improved ID3 algorithm was enhanced. Finally, it is proved that the improved algorithm is more efficient and faster than the original method with an example. The algorithm is validated by an implementation.

关 键 词：数据挖掘 决策树 粗糙集 ID3 属性重要度 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]