CAJ | 학술논문

为深入研究和评估蚁群算法在分类规则挖掘应用中具有的特点和作用，针对目前基本蚁群算法在数据挖掘方面所存在的不足，引入了改进的蚁群算法模型最大最小蚂蚁系统( MMAS)。并根据分类算法比较原则，通过实验分析对分类规则挖掘算法进行比较。根据使用不同数据集实验结果的对比分析，从仿真的精确度、速度等方面展示和证实了基于改进的蚁群算法模型MMAS的数据分类规则挖掘工具AntMiner+在分类规则挖掘中体现出的特点和优势。
위심입연구화평고의군산법재분류규칙알굴응용중구유적특점화작용，침대목전기본의군산법재수거알굴방면소존재적불족，인입료개진적의군산법모형최대최소마의계통( MMAS)。병근거분류산법비교원칙，통과실험분석대분류규칙알굴산법진행비교。근거사용불동수거집실험결과적대비분석，종방진적정학도、속도등방면전시화증실료기우개진적의군산법모형MMAS적수거분류규칙알굴공구AntMiner+재분류규칙알굴중체현출적특점화우세。
In order to study and evaluate the features and functions of ant colony algorithm in classification rule mining applications in-depth,aiming at the deficiencies of the basic ant colony algorithm,introduce improved ant colony algorithms,the Max-Min Ant System (MMAS). And according to the comparison principle of the classification rule mining algorithm,make a comparison for classification rule mining algorithms through the experimental analysis. The results show that the AntMiner+based on MMAS model has great advanta-ges in the classification rule mining from simulation accuracy and speed.