CAJ | 학술논문

本文提出了一种多蚁型的蚁群模型和在此模型上的一个具有移动空间的蚁群聚类算法。每个类簇都带有一个移动空间，使同类的蚂蚁紧密地聚在一起。移动空间为每个类簇选出一个代表蚂蚁用来长距离的类簇移动。代表蚂蚁通过连接将类簇中不相似的大量数据移动到与其相似的代表蚂蚁处，减少了蚂蚁的搜索时间，提高聚类性能。针对UCI数据集进行实验，结果表明新算法提高了聚类效果。
본문제출료일충다의형적의군모형화재차모형상적일개구유이동공간적의군취류산법。매개류족도대유일개이동공간，사동류적마의긴밀지취재일기。이동공간위매개류족선출일개대표마의용래장거리적류족이동。대표마의통과련접장류족중불상사적대량수거이동도여기상사적대표마의처，감소료마의적수색시간，제고취류성능。침대UCI수거집진행실험，결과표명신산법제고료취류효과。
The paper presents an ant colony model based on multiple heterogeneous ants and an ant colony clustering algorithm with movement zones. Here ants maintain a movement zone around each cluster, keeping ants close to data items. Each movement zone elect the representatives that are responsible for all long distance movement. The representatives can, probabilistically, pass a large number of dissimilar objects it has to any other similar representative, greatly reducing cluster merging time and improving the performance of clustering. Experimental results show that the new algorithm can improve the quality of clustering.