核属性蚁群算法的规则获取

蚁群算法是一种新型的模拟进化算法,研究已经表明该算法具有许多优良的性质,并且在优化计算中已得到了很多应用.粗糙集理论作为一种智能数据分析和数据挖掘的新的数学工具,其主要优点在于它不需要任何关于被处理数据的先验或额外知识.本文从规则获取和优化两方面研究基于粗糙集理论和蚁群算法的分类规则挖掘方法.通过研究决策表和决策规则系数,建立基于粗糙集表示和度量的知识理论,将粗糙集理论与蚁群算法融合,采用粗糙集理论进行属性约简,利用蚁群算法获取最优分类规则,优势互补.实验结果比较表明,算法获取的分类规则,具有良好的预测能力和更为简洁的表示形式.

Core Attribute and Ant Colony Algorithm for Acquisition of Rules

Ant colony algorithm is a novel simulated evolutionary algorithm which shows many promising characters and it is numerous applied to optimization computation.Rough set theory is a new mathematical tool for intelligent data analysis and data mining,the main advantage of which is it doesn't require any prior or additional knowledge about the data.The algorithm in this paper is based on the combination of rough set theory and ant colony algorithm for mining classification rules.The method is studied in tow aspects:decision rules acquisition and optimization.It constructs knowledge theory based on rough set denotation and measure according to coefficients of decision rule and decision table.First,it adopts rough set theory to attribute reduction.Second,it makes use of the ant colony algorithm to acquisitive rules.Finally,it develops enough advantage of the tow methods.Experimental results show that the algorithm can discover better classification rule with better predictive accuracy and simpler rule form.