摘要
通过分析分类规则与训练集之间的映射关系,采用集合的相关运算寻找特征规则及相应特征集,从而消除分类规则集中存在的冗余,并在此基础上提出了基于集合运算的分类规则处理算法(PASO)。最后,以恒星光谱数据为背景,实验验证了该方法的正确性和可行性。
In this paper, a processing algorithm of classification rule based on set operation is presented by analyzing the mapping relationship between rules and training data, and set operations are used to get characteristic rule and its characteristic set, so that the redundant rules may be discovered and eliminated. In the end, experimental results validate its correctness and efficiency by using star spectra data as the decision system.
出处
《太原科技大学学报》
2011年第2期131-134,共4页
Journal of Taiyuan University of Science and Technology
基金
太原科技大学青年科学基金(20083015)
关键词
数据挖掘
分类规则
集合
恒星光谱数据
DM （Data Mining） ,classification rule ,set ,star spectra data
