A Comparative Study of Efficient Initialization Methods for the K-Means Clustering Algorithm

Abstract

K-means is undoubtedly the most widely used partitional clustering algorithm. Unfortunately, due to its gradient descent nature, this algorithm is highly sensitive to the initial placement of the cluster centers. Numerous initialization methods have been proposed to address this problem. In this paper, we first present an overview of these methods with an emphasis on their computational efficiency. We then compare eight commonly used linear time complexity initialization methods on a large and diverse collection of data sets using various performance criteria. Finally, we analyze the experimental results using non-parametric statistical tests and provide recommendations for practitioners. We demonstrate that popular initialization methods often perform poorly and that there are in fact strong alternatives to these methods.

17 pages, 1 figure, 7 tables

References in corpus (1)

• Improving the Performance of K-Means for Color Quantization

Cited by in corpus (4)

• Deterministic Initialization of the K-Means Algorithm Using Hierarchical Clustering
• Effective Deterministic Initialization for $k$-Means-Like Methods via Local Density Peaks Searching
• On improved bound for measure of cluster structure in compact metric spaces
• Interdependence of clusters measures and distance distribution in compact metric spaces