摘要
不完全信息系统中遗失数据的补充和规则的提取,一直是数据挖掘技术面临的重要问题。文中给出了一种基于决策树来求解此问题的算法。对于给定的不完全决策表,该算法应用改进的ID3算法来构造决策树,在构造决策树的过程中对遗失值进行补充。对于不能在决策树上补充的遗失值,定义了一种相关对象之间的相似度来填充。该算法简单,易于操作。
Missing data filling and rules extraction in incomplete decision table are two important data mining problems. Based on decision tree, the paper gives an algorithm to solve these problems. For a given incomplete decision table, the algorithm constructs decision tree using the improved ID3 algorithm, and fills the missing data in the process of constructing the decision tree. A similar measure to fill the missing data that cant be filled in the process of constructing the decision tree is defined. The algorithm is simple and easily handled. The algorithm is illuminated with an example.
出处
《计算机应用》
CSCD
北大核心
2003年第11期17-19,22,共4页
journal of Computer Applications
基金
国家 97 3规划资助项目 (G1 9980 3 0 6 )
关键词
不完全决策表
遗失值
数据补充
决策树
规则提取
incomplete decision table
missing data
data filling
decision tree
rule extraction
