摘要
提出了一种基于粗糙集理论的最简规则挖掘方法，它是一个采用基于分类正确度的粗糙集模型进行多概念分类规则挖掘的新方法，能有效处理决策表的不一致性，采用启发式算法，挖掘出满足给定精确度的最简产生式规则知识。用多个UCI数据集对算法进行了测试，并且与著名的Rosetta软件进行实验对比，结果说明此方法大大提高了总的数据约简量，可以有效地简化最终得到的规则知识。
The paper presents a data mining approach based on rough set theory, which is used to mine multi-concept classification rules based on the variable precision rough set model. It deals with inconsistent examples, uses heuristic algorithm to build concise production rules for each concept satisfying the given certainty factor. The paper also uses many UCI data sets to test the proposed approach and compares with the Rossetta tool. The results show this method greatly improves the total data reduction and efficiently simplifies the rule sets.
出处
《计算机工程》
CAS
CSCD
北大核心
2003年第20期77-79,共3页
Computer Engineering
基金
教育部科学技术研究项目 (03102)
关键词
数据挖掘
粗糙集
约简
不一致例子学习
Data mining
Rough set
Reduction
Learning from inconsistent examples
