k-means++

Versione 1.7.0.0 (1,74 KB) da Laurent S
Cluster multivariate data using the k-means++ algorithm.
4.7
(22)
14,3K download
Aggiornato 11 feb 2013

Visualizza la licenza

An efficient implementation of the k-means++ algorithm for clustering multivariate data. It has been shown that this algorithm has an upper bound for the expected value of the total intra-cluster distance which is log(k) competitive. Additionally, k-means++ usually converges in far fewer than vanilla k-means.

Cita come

Laurent S (2026). k-means++ (https://it.mathworks.com/matlabcentral/fileexchange/28804-k-means), MATLAB Central File Exchange. Recuperato .

Compatibilità della release di MATLAB
Creato con R2012b
Compatibile con qualsiasi release
Compatibilità della piattaforma
Windows macOS Linux
Categorie
Scopri di più su Statistics and Machine Learning Toolbox in Help Center e MATLAB Answers
Tag Aggiungi tag
Riconoscimenti

Ispirato da: Kmeans Clustering

Ispirato: kmeans_varpar(X,k), Sparsified K-Means

Versione Pubblicato Note della release
1.7.0.0

Fixed bug with 1D datasets (thanks Xiaobo Li).
1.6.0.0

Improved handling of overclustering (thanks Sid S) and added a screenshot.
1.5.0.0

Small bugfix.
1.4.0.0

Removed dependency on randi for R2008a or lower (thanks Cassie).
1.3.0.0

Even faster, even less code and also fixed a few small bugs.
1.0.0.0