An optimized method for selection of the initial centers of k-means clustering

Authors:
YongJun Zhang;Enxiu Cheng
Affiliations:
Shandong Institute of commerce and technology, Jinan, China,National Engineering Research Center for Agricultural Products Logistics, Jinan, China;National Engineering Research Center for Agricultural Products Logistics, Jinan, China
Venue:
IUKM'13 Proceedings of the 2013 international conference on Integrated Uncertainty in Knowledge Modelling and Decision Making
Year:
2013

Citing 3
Cited 0

Automated Variable Weighting in k-Means Type Clustering

IEEE Transactions on Pattern Analysis and Machine Intelligence
A simple and fast algorithm for K-medoids clustering

Expert Systems with Applications: An International Journal
Data Mining: Concepts and Techniques

Data Mining: Concepts and Techniques

Quantified Score

Hi-index	0.00

Visualization

Abstract

At present, K-Means algorithm as a method of clustering based on partition has more applications. By analyzing the problem of K-Means, we propose a kind of optimized algorithm to select initial clustering points, which utilize an adjacent similar degree between points to get similarity density, then by means of max density points selecting our method heuristically generate clustering initial centers to get more reasonable clustering results. This method compares to traditional methods lie in it will get suitable initial clustering centers and also has more stable clustering result. Finally, through comparative experiments we prove the effectiveness and feasibility of this algorithm.