第25卷第8期 - 计算机应用.pdf

第25卷第8期 计算机应用 V01．25 No．8 2005年8月 Computer Applications Aug．2005 文章编号：1001—9081(2005)08—1853—03 基于值约简和决策树的最简规则提取算法 罗秋瑾，陈世联 (昆明理工大学理学院，云南昆明650093) (qiujinluo@sina．eom) 摘 要：粗糙集理论中的值约简和数据挖掘领域中的决策树都是有效的分类方法，但二者都有其 局限性。将这两种方法结合起来，生成一种新的基于值核的极小化方法对决策树进行修剪，提出了约 简规则的判定准则，缩小了约简的范围，最后再对生成的规则进行极大化处理，以保证规则覆盖信息 的一致性，实验验证了该算法的有效性。 关键词：粗糙集；数据挖掘；决策树；值约筒；分类规则 中图分类号：TP311．13 文献标识码：A Algorithm based on value reduction and decision tree to generate minimal rules LU0 Oiu-jin．CHEN Shi—lian (School ofScience,Kunming University ofScience and Technology,Kunming Yunnan 650093，China) Abstract：Value reduction in rough set theory and decision tree in data mining are effectively used in the classification ， but each of them has shoacomings．Those two methods were combined to generate 8 new minim【al method based on value core to pollard the decision tree．Then judgmental standards of rule reduction were proposed to decrease the quantity of reducted rules．In the end，the classing rules were dealed with by maximal method to ensure the consistency of the knowledge contained by the rules．The algorithm is efficient which Was proved by the experiment． Key words：rough set；data mining；decision tree；value reduction；classing rule 粗糙集中的值约简、神经网络、统计模型、贝叶斯分类器和决 进行极大化处理 。 奏 的 竺是 ． I， 竺窭 1．1基于粗糙集的值约简 冀 树 贝 墨它引 莩 统