基于网格的快速搜寻密度峰值的聚类算法优化研究

CFSFDP是基于密度的新型聚类算法，可聚类非球形数据集，具有聚类速度快、实现简单等优点。然而该算法在指定全局密度阈值dc时未考虑数据空间分布特性，导致聚类质量下降，且无法对多密度峰值的数据集准确聚类。针对以上缺点，提出基于网格分区的CFSFDP（简称GbCFSFDP）聚类算法。该算法利用网格分区方法将数据集进行分区，并对各分区进行局部聚类，避免使用全局dc，然后进行子类合并，实现对数据密度与类间距分布不均匀及多密度峰值的数据集准确聚类。两个典型数据集的仿真实验表明，GbCFSFDP算法比CFSFDP算法具有更加精确的聚类效果。...

Full description

Saved in:
Bibliographic Details
Published in计算机工程与科学 Vol. 39; no. 5; pp. 964 - 970
Main Author 孙昊 张明新 戴娇 尚赵伟
Format Journal Article
LanguageChinese
Published 苏州大学计算机科学与技术学院,江苏苏州,215006%常熟理工学院计算机科学与工程学院,江苏常熟,215500%重庆大学计算机学院,重庆,400030 2017
Subjects
密度阈值
类合并
网格分区
聚类
聚类
类合并
density threshold
密度阈值
clustering
网格分区
grid partition
merging clusters
Online AccessGet full text
ISSN1007-130X
DOI10.3969/j.issn.1007-130X.2017.05.022

Cover

More Information
Summary:CFSFDP是基于密度的新型聚类算法，可聚类非球形数据集，具有聚类速度快、实现简单等优点。然而该算法在指定全局密度阈值dc时未考虑数据空间分布特性，导致聚类质量下降，且无法对多密度峰值的数据集准确聚类。针对以上缺点，提出基于网格分区的CFSFDP（简称GbCFSFDP）聚类算法。该算法利用网格分区方法将数据集进行分区，并对各分区进行局部聚类，避免使用全局dc，然后进行子类合并，实现对数据密度与类间距分布不均匀及多密度峰值的数据集准确聚类。两个典型数据集的仿真实验表明，GbCFSFDP算法比CFSFDP算法具有更加精确的聚类效果。
Bibliography:43-1258/TP
clustering ; density threshold ; grid partition;merging clusters
The CFSFDP is a clustering algorithm based on density peaks, which can cluster arbitrary shape data sets, and has the advantages of fast clustering and simple realization. However, the global density threshold de, which can lead to the decrease of clustering quality, is specified without the consid- eration of spatial distribution of the data. Moreover, the data sets with multi-density peaks cannot be clustered accurately. To resolve the above shortcomings, we propose an optimized CFSFDP algorithm based on grid （GbCFSFDP）. To avoid the using of global d,, the algorithm divides the data sets into smaller partitions by using the grid partitioning method and performs local clustering on them. Then the GbCFSFDP merges the sub classes. Data sets, which are unevenly distributed and have multi-density peaks, are correctly classified. Simulation experiments of two typical data sets show that the GbCFS- FDP algorithm is more accurate than the CF
ISSN:1007-130X
DOI:10.3969/j.issn.1007-130X.2017.05.022