加权三支决策增量软聚类算法及性能分析

现有的增量聚类算法虽然解决了数据增量和类簇重叠问题，但在距离度量时没有考虑属性重要度不同，且普遍拥有较高的时间复杂度。针对以上问题，提出一种基于属性重要度的加权三支决策增量软聚类算法（W-TIOC-TWD算法），将属性重要度考虑到距离度量中，弥补了现有算法在聚类过程中将所有属性的重要程度视为相等的不足。该算法还引入离群点概念，降低了算法的时间复杂度。基于人工数据集和UCI数据集的实验结果表明，W-TIOC-TWD算法的聚类准确率优于比较算法。

A Weighted Three-way Decision Incremental Clustering Algorithm and Performance Analysis

Though the existing incremental clustering algorithms can solve the problem of data increment and class overlap， those algorithms do not consider the difference of attribute importance in distance measurement and generally have a higher time complexity. To solve the above problems， this paper proposes the W-TIOC-TWD algorithm. Taking attribute importance into the calculation of distance measure， this algorithm can cover the shortage that equally regard the importance of all attributes in the process of clustering. Moreover， the definition of outlier point is proposed， which improves the time efficiency of this algorithm. To verify the effectiveness and accuracy of this algorithm， experiments on artificial datasets and UCI datasets are implemented. Experimental results show that the W-TIOC-TWD algorithm outperform the comparison algorithms.