对层次聚类算法的改进

Improving Hierarchical Clustering Algorithm

介绍了层次聚类算法的基本思想,在原算法的基础上,通过对类间距离进行预排序从而避免合并类时重新计算类间距离,该算法降低了计算的复杂度.实验表明,改进后的算法在与原算法聚类效果相同的情况下运行速度明显提高.

The agglomerative hierarchical clustering algorithm is introduced in this paper based on the standard algorithm. Through sorting the distance between clusters to avoid recomputing the distance between clusters we proposed the improved algorithm,which decreases the complexity,According to the experiment, the improved algorithm has the same clustering results as the standard agglomerative hierarchical clustering algorithm,but the speed is higher.