摘要
以概念格为分类模型,引入知识合并思想,并针对大规模数据的分类求解以及过拟合问题引入剪枝策略,从而得到分类剪枝概念格模型,在此基础上提出了基于分布式概念格模型的强分类规则提取算法;通过理论证明了算法的正确性,并通过实验证明了算法的可行性。
The algorithm of mining strong classification rules based on the distributed extended concept lattice（ECL） is presented, in which the concept lattice is used as the classification model and the method of knowledge combination is adopted. In order to mine strong classification rules in large scale databases, the mechanism of pruning is imported to handle the problem of overfitting. The correctness of the algorithm is proved by theory and its feasibility is shown by experiment.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
北大核心
2007年第2期132-136,共5页
Journal of Hefei University of Technology：Natural Science
基金
安徽省自然科学基金资助项目(050420207)
合肥工业大学科研发展基金资助项目(050504F)
关键词
分类规则
分布式数据挖掘
概念格
过拟合
剪枝
classification rule
distributed data mining
concept lattice
overfitting
pruning
