【摘要】 为了提高基于网格聚类技术的聚类精度和效率,提出一种新的基于网格距离的高精度聚类算法。该算法一方面通过参考网格在逻辑空间的相对距离进行聚类,从而弥补了大多数计算网格之间距离的算法中需要大量数学运算的不足,另一方面,提出了一种新的边界点处理技术。用实际数据集进行的,实验结果表明,该技术能够有效地提取有意义的边界点,运行速度快、聚类精度高。更多还原

【Abstract】 In order to improve the precision and efficiency of the grid-based clustering technology,in the paper it presents a new high-precision clustering algorithm based on distance between grids.On one hand,the algorithm deals with datasets by referring relative distances between grids in logic space,which makes up the deficiency of some algorithms that need much mathematical operation to support.On the other hand,it presents a new technique to deal with boundary points of clusters.Besides,the experiment has been carried out with real datasets,the result shows that the algorithm is efficient in picking up the meaningful boundary points with precise clustering and fast running time.更多还原