摘要
K-means算法需要人工设定聚类个数且易受孤立点影响,根据这个缺陷提出了一种新的改进算法。改进算法通过设定初始值及初始值的最大值,在聚类过程中自动获取聚类数k。实验结果表明,该算法在一定程度上缓解了K-means算法对初始值敏感及受孤立点影响的问题,能产生高质量的聚类结果。
Considering the vital need for clustering number manally set and the vulnerable defects by isolated point of K-means,a new improved algorithm is proposed.The improved algorithm can acquire the number of cluster k automatically in the clustering process through setting the initial value and the maximum of the initial value.The experimental results indicate that the algorithm,to some extent,can alleviate the K-means algorithm which is sensitive to the initial value and subject to the impact of the isolated point and can produce high-quality clustering results.
出处
《佛山科学技术学院学报（自然科学版）》
CAS
2010年第2期49-52,共4页
Journal of Foshan University(Natural Science Edition)
