k-means++ clustering

The k-means algorithm is one of the most commonly used clustering methods. The algorithm strives for partitioning into clusters minimizing the sum of intracluster distances squared. The base method is fully detailed in Wikipedia. The ALGLIB package implements an algorithm version that is called "k-means++".

Assumptions

• The number of clusters is predetermined
• The Euclidean distance is an adequate measure of similarity of the set elements

Manual entries

This article is intended for personal use only.

Download ALGLIB