ClusterTools Contents

ClusterTools User Guide



Mean shift mode-seeking clustering


 A Dataset of M objects (rows).
 K Vector with desired numbers of clusters. Default is a set of N clusterings with numbers that naturally arise from the data.
 MSIZE Number of objects (M) above which the dataset is preclustered by
 CLUSTM, reducing it to MSIZE objects. Default MSIZE = 5000. Use
 MSIZE = inf to avoid perclustering.

 LAB M*N array with the results of the multi-level clusterings for the
 M objects. The columns refer to the N clusterings. They yield for  the objects the prototype indices of the clusters they belong to.


This is a wrapper around MEANSHIFT, which is a wrapper around  MeanShiftCluster routine by Bryan Feldman and Bart Finkston. These  routines do not return clusterings with a preset number of clusters K.  The resulting multilevel clustering is first made nested by RECLUSTN.  After that the desired number of clusters is realized by RECLUSTK.


 randreset;                     % take care of reproducability
 data = gendatclust1(20000);    % generate 20000 objects in 10 clusters
                                % Run Mean Shift clustering
 lab = clusts(data,[2 5 10 18 30 50 100],2000);
                                % Show scatterplot for 10 clusters
 figure; scatn(lab(:,3),data,'Mean Shift'); 
 figure; clusteval(lab,data);   % Evaluation by active learning


Cheng, Y. "Mean shift, mode Seeking, and clustering", IEEE Transactions on PAMI, vol. 17, no. 8, pp. 790-799, 1995.

R.P.W. Duin, A.L.N. Fred, M. Loog, and E. Pekalska, Mode Seeking Clustering by KNN and Mean Shift Evaluated, Proc. SSPR & SPR 2012, LNCS, vol. 7626, Springer, 2012, 51-59.

See also

datasets, mappings, cluste, clustf, clusth, clustk, clustkh, clustm, reclustn, preclust, clusteval, clustcerr, clustc,

ClusterTools Contents

ClusterTools User Guide

This file has been automatically generated. If badly readable, use the help-command in Matlab.