一种基于网格的等密度线聚类算法

摘要

We proposed a grid-based density-isoline clustering algorithm.The algorithm meshes the space where the samples in,with the thought of density distribution,and it can automatically discover any kind of shape,and has great time complexity and space complexity.In addition,we achieved the calculation of the sample space of different types of volume.It has good clustering effect and volume grid-based computing capabilities.%提出了一种基于网格的等密度线聚类算法,通过对样本所在空间进行网格划分,从样本分布等密度线图的思想出发,可自动发现任何形状的类,时间复杂度和空间复杂度较好,实现了对不同类样本空间容积的计算,具有很好的聚类效果和容积计算能力.