基于免疫算法的粗糙集知识约简

Rough Set Knowledge Reduction Based on Immune Algorithm

在基于粗糙集理论的知识发现过程中,减小属性约简复杂度问题是重要研究内容之一,是在保持信息系统分类能力不变的基础上,删除冗余知识。通过在知识表达系统中的决策属性支持度来描述由条件属性所提供的知识对整体决策的支持程度,通过相对重要程度来描述条件属性对决策属性的重要性。然后利用免疫网络机理和约简算法融合,构造免疫网络约简算法,把相对核加入初始种群加快收敛速度。最后,以经典的实例分析表明,该方法是求解知识约简问题的快速有效方法。

Knowledge reduction is one of the important issues of Knowledge Discovery based on rough set theory, and it is to remove superfluous knowledge from information systems while preserving the consistency of classifications. The support degree of the knowledge supplied by condition attribute for the whole decision was described by the decision attribute support degree applied in knowledge express system, and relative importance degree was described . Then , the hybrid Immune network reduction Algorithm is proposed , and the relative core is added to the initial popular to improve the convergence speed. At last, an example proves that the algorithm is effective.