[data analysis] data clustering and grouping based on kmeans, including Matlab source code

1 brief introduction

From theory , Design program and code implementation , Explain how data mining K- Mean clustering algorithm , utilize Matlab The flexible programming function of . Illustrated by experimental teaching practice , In order to cultivate innovative thinking and practical ability throughout experimental teaching , utilize Matlab Simulation K- Mean clustering has better practical significance .

Clustering is the process of recombining data sets into multiple classes according to the similarity of data . Clustering is different from classification : Classification is a supervised learning method , The characteristics of data classes are known before classification ; Clustering is an unsupervised learning method , Do not know the characteristics of data classes before clustering . In the experimental teaching of data mining , Teachers often only pay attention to the clustering algorithm itself , Not really understand the essence of clustering , Only the clustering algorithm is written in a fixed form , Even only the program that cannot be changed and the pre-set divided array are clustered , Misunderstood the essence of clustering , The flexible algorithm and clustering idea are not integrated into the experiment . The purpose of cultivating students' innovative thinking and ability in the experiment has not been achieved .

2 Part of the code

function []=show(X,cidx,ctrs,varargin)%{
   show[X,cidx,ctrs]       X A matrix of input raw data ,cidx,ctrs by k_meansnD The result returned ,                         They are n Grouping and centroid coordinates of iterations .show[X,cidx,ctrs,stp]    The number of iterations displayed is stp The grouping at the time .​%}if      nargin<3        error('there is not enough input arguments >.<#~ ')  elseif  nargin==3        idx=cidx(:,end);    % If the input variable is 3 individual ,idx Fu Cheng cidx The last column of , That is, the final grouping .        ctr=ctrs(:,:,end);  %ctr Be endowed with ctrs The last page of , That is, the final center of mass coordinates .elseif  nargin==4        stp=varargin{1};        idx=cidx(:,stp);        ctr=ctrs(:,:,stp);elseif  nargin>4        error('there is too many input arguments >.<#~ ')    end     %end ifk=size(ctrs,1);C=jet(k);d=size(X,2);if      d==2        for i=1:k        plot(X(idx==i,1),X(idx==i,2),'.','color',C(i,:));        hold on;        plot(ctr(i,1),ctr(i,2),'kx')        hold on;        end %end for        hold off;elseif  d==3        for i=1:k        plot3(X(idx==i,1),X(idx==i,2),X(idx==i,3),'.','color',C(i,:));        hold on;        plot3(ctr(i,1),ctr(i,2),ctr(i,3),'kx')        hold on;        end %end for        hold off;end     %end ifgrid onend     %end function​

3 Simulation results

4 reference

About bloggers ： Good at intelligent optimization algorithms 、 Neural networks predict 、 signal processing 、 Cellular automata 、 The image processing 、 Path planning 、 UAV and other fields Matlab Simulation , relevant matlab Code problems can be exchanged by private letter .

Some theories cite network literature , If there is infringement, contact the blogger to delete .