CAJ | 학술논문

针对K—Modes算法的不足，提出了一种基于信任值的分类属性聚类算法TrustCCluster，该算法不需预先给定聚类个数，聚类结果稳定且不依赖于初始值的选取。在真实数据上验证了TrustC—Cluster聚类算法，并与K—Modes和P—Modes算法进行了对比，实验结果表明TmstCCluster算法是有效、可行的。
침대K—Modes산법적불족，제출료일충기우신임치적분류속성취류산법TrustCCluster，해산법불수예선급정취류개수，취류결과은정차불의뢰우초시치적선취。재진실수거상험증료TrustC—Cluster취류산법，병여K—Modes화P—Modes산법진행료대비，실험결과표명TmstCCluster산법시유효、가행적。
For the shortage of K-Modes algorithm, a categorical attribute clustering algorithm TrustCCluster based on trust val-ue is proposed, the algorithm does not need to pre-specify the number of clusters, and clustering results do not depend on the se-lection of the initial values. TrustCCluster clustering algorithm is verified on the real data, and compared with the K-Mode and P-Modes algorithms, the result shows that TrustCCluster algorithm is feasible and effective.