基于属性约简的恒星光谱数据分类规则挖掘系统研究

Research on Classification Rules Mining System about Stellar Spectrum Data Based on Attribute Reduction

知识约简与决策规则的提取是粗糙集理论研究的核心内容。本文针对新加入对象相对于原来的极小决策算法而言是全新的这一情况,提出了一种基于粗糙逻辑的增量式属性约简算法,从而避免每次从庞大的原始决策表开始约简,提高了效率。在此基础上,采用VC^(++)和Oracle9i为开发工具,设计与实现了基于属性约简的恒星光谱数据分类规则挖掘系统,从而为实现恒星光谱数据的自动分类提供了一种有效途径。

Knowledge reduction and the extraction of decision rules are very important in the rough set theory. In this paper, a dynamic algorithm of attribute reduction based on rough logic is presented. The algorithm can avoid reduction from large original decision table when new object is added, and improve the efficiency of attribute reduction. According to this algorithm, the system of classification rules about stellar spectrum data based on attribute reduction is developed through using VC^(++) and Oracle9i. It can afford an effective method to auto-classification of stellar spectrum data.