Abstract: K-means clustering algorithm is a commonly algorithm applied in data mining and data analysis, but it has the disadvantages of relying on the initial value and easy to fall into the local optimum. For these shortcomings, this paper proposes an improved K-means clustering which is optimized by the lightning attachment procedure Optimization (LAPO) , which overcomes the difficulty of selecting the initial value of the clustering algorithm. This optimum improves the accuracy of the K-means clustering algorithm, and reduces the possibility of falling into a local optimum. Six real data sets are selected from the UCI data set for simulation experiments. The results show that the improved clustering algorithm has better accuracy and robustness.

Key words: clustering, lightning attachment procedure optimization, data processing, K-means

摘要： K-means聚类算法是在数据挖掘和数据分析中一种常用算法,但是其存在依赖初始值和易陷入局部最优值的缺陷,针对这些不足,本文提出一种闪电分叉过程算法优化的K-means聚类,克服聚类算法在初始值选择困难的问题,提高K-means聚类算法的求解精度,降低陷入局部最优的可能性。从UCI数据集中选取6个真实的数据集进行仿真实验,结果表明本文改进后的聚类算法有更好的求解精度和鲁棒性。

关键词: 聚类, 闪电分叉过程算法, 数据处理, K-均值聚类