Study on Density Parameter and Center-Replacement Combined K-means and New Clustering Validity Index

Clustering is a classic data mining technology, which has been widely used in pattern recognition, machine learning, artificial intelligence and other fields. Through clustering analysis, the deep structure of the target data set can be effectively discovered. As a commonly used partitioning and clu...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Ji suan ji ke xue Jg. 49; H. 1; S. 121 - 132
Hauptverfasser: Zhang, Ya-di, Sun, Yue, Liu, Feng, Zhu, Er-zhou
Format: Journal Article
Sprache:Chinesisch
Veröffentlicht: Chongqing Guojia Kexue Jishu Bu 01.01.2022
Editorial office of Computer Science
Schlagworte:
Algorithms
Artificial intelligence
Cluster analysis
Clustering
clustering algorithm|clustering validity index|optimal clustering number|cluster center|data mining
Data mining
Datasets
Density
Machine learning
Outliers (statistics)
Parameters
Pattern recognition
ISSN:1002-137X
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Clustering is a classic data mining technology, which has been widely used in pattern recognition, machine learning, artificial intelligence and other fields. Through clustering analysis, the deep structure of the target data set can be effectively discovered. As a commonly used partitioning and clustering algorithm, K-means has the advantages of being simple to implement and capable of handling large data. However, affected by the convergence rules, the K-means algorithm is still very sensitive to the selection of initial cluster centers and cannot be It handles non-convex distributions and data sets with outliers well. An improved K-means algorithm DC-Kmeans based on density parameters and center replacement is proposed in this paper. The algorithm uses the density parameters of the data object to gradually Determine the initial cluster center, and use the center replacement method to update the initial center that deviates from the actual position, so it is more accurate than the traditional clustering alg
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1002-137X
DOI:10.11896/jsjkx.201100148