A Cluster Algorithm Identifying the Clustering Structure
Zhiwei Sun
Abstract
Zhiwei Sun
Abstract
Cluster analysis is a primary method for database mining. Most of clustering algorithms require input parameters which are hard to determine but have a significant influence on the clustering result. Furthermore, for many real-datasets there does not exist a global parameter setting for which the result of the clustering algorithm describes the intrinsic clustering structure accurately. We introduce a new algorithm which produces a clustering explicitly. The algorithm first gets the approximate density of every point using the grid, and then uses k-means algorithm to get the boundary of cluster structure with the data of point density, at last it uses values of boundary as the parameters of the next step which can get the finical cluster result. Both theory analysis and experimental results confirm CluICS can cluster data of varying density with automatic setting different parameters in different partitions and its efficiency is much higher than DBSCAN algorithm.
OpenAlex reports 3 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Cluster analysis is a primary method for database mining. Most of clustering algorithms require input parameters which are hard to determine but have a significant influence on the clustering result. Furthermore, for many real-datasets there does not exist a global parameter setting for which the result of the clustering algorithm describes the intrinsic clustering structure accurately. We introduce a new algorithm which produces a clustering explicitly. The algorithm first gets the approximate density of every point using the grid, and then uses k-means algorithm to get the boundary of cluster structure with the data of point density, at last it uses values of boundary as the parameters of the next step which can get the finical cluster result. Both theory analysis and experimental results confirm CluICS can cluster data of varying density with automatic setting different parameters in different partitions and its efficiency is much higher than DBSCAN algorithm.
Key concepts: Cluster analysis, DBSCAN, Computer science, CURE data clustering algorithm, Correlation clustering, Data mining, Cluster (spacecraft), k-medians clustering