模糊聚類函數(shù)的matlab代碼

1、模糊聚類函數(shù)的matlab代碼functionR=coesim1(datin,m)%聚類分析相關(guān)系數(shù)計算公式1數(shù)量積%R=coesim1(datin,C)%m是乘積和最大值的一個參數(shù)，取決于輸入數(shù)據(jù)datinerror(nargchk(2,2,nargin)NM=size(datin);R=zeros(N,N);fori=1:Nforj=1:Nfork=1:MR(i,j)=R(i,j)+(datin(i,k)*datin(j,k)/m;endendR(i,i)=1;endfunctionR=coesim2(datin)%聚類分析相關(guān)系數(shù)計算公式2夾角余弦%R=coesim2(datin)err

2、or(nargchk(1,1,nargin)NM=size(datin);R=zeros(N,N);t=zeros(1,3);fori=1:Nforj=1:Nfork=1:Mt(1)=t(1)+datin(i,k)*datin(j,k);t(2)=t(2)+datin(i,k)7;t(3)=t(3)+datin(j,k)7;endR(i,j)=abs(t(1)/sqrt(t(2)*t(3);t=zeros(1,3);endendfunctionR=coesim3(datin)%聚類分析相關(guān)系數(shù)計算公式3相關(guān)系數(shù)%R=coesim3(datin)error(nargchk(1,1,nargin)

3、NM=size(datin);R=zeros(N,N);datinmean=mean(datin);t=zeros(1,3);fori=1:Nforj=1:Nfork=1:Mt(1)=t(1)+abs(datin(i,k)-datinmean(i)*(datin(j,k)-datinmean(j);t(2)=t(2)+(datin(i,k)-datinmean(i)F2;t(3)=t(3)+(datin(j,k)-datinmean(j)A2;endR(i,j)=t(1)/sqrt(t(2)*t(3);t=zeros(1,3);endendn+=XJxX-中大于。的元素個數(shù)n-為上面那組數(shù)中小

4、于0的元素個數(shù)口匸貝十_n_functionR=coesim5(datin)%聚類分析相關(guān)系數(shù)計算公式5非參數(shù)法%R=coesim5(datin)error(nargchk(1,1,nargin)NM=size(datin);R=zeros(N,N);datinmean=mean(datin);t=zeros(1,2);fori=1:Nforj=1:NTi=zeros(1,M);Tj=zeros(1,M);Ti=datin(i,:)-datinmean(i);Tj=datin(j,:)-datinmean(j);t(1)=length(find(Ti.*Tj0);t(2)=length(fin

5、d(Ti.*Tj0);R(i,j)=abs(t(1)-t(2)/(t(1)+t(2);endend_E芝二工阪說伍低，XjQMSK=imax(xik-x)k)functionR=coesim6(datin)%聚類分析相關(guān)系數(shù)計算公式6最大最小法%R=coesim6(datin)error(nargchk(1,1,nargin)NM=size(datin);R=zeros(N,N);fori=1:Nforj=1:NR(i,j)=sum(min(datin(i,:),datin(j,:)/.sum(max(datin(i,:),datin(j,:);endend_E旻二：Lmhi絢對XjQ昇l(x

6、dXjic)functionR=coesim7(datin)%聚類分析相關(guān)系數(shù)計算公式7算術(shù)平均最小法%R=coesim7(datin)error(nargchk(1,1,nargin)NM=size(datin);R=zeros(N,N);fori=1:Nforj=1:NR(i,j)=sum(min(datin(i,:),datin(j,:)/.(sum(sum(datin(i,:);datin(j,:)/2);endendfunctionR=coesim9(datin)%聚類分析相關(guān)系數(shù)計算公式9絕對值指數(shù)法%R=coesim9(datin)error(nargchk(1,1,nargin

7、)NM=size(datin);R=zeros(N,N);fori=1:Nforj=1:NR(i,j)=exp(-sum(abs(datin(i,:)-datin(j,:);endendend(1當(dāng)i二j(1一CSK=l|ik一Xjk|當(dāng)i工jfunctionR=coesim11(datin,C)%聚類分析相關(guān)系數(shù)計算公式11絕對值減數(shù)法%R=coesim11(datin)%r(i,j)=1,wheni=j;%=1-C*(M/k=1)|Xik-Xjk|,whenj!=i;error(nargchk(2,2,nargin)if(C1),error(Cmustbetween0and1);endNM

8、=size(datin);R=ones(N,N);fori=1:Nforj=1:Nfork=1:MR(i,j)=R(i,j)-C*abs(datin(i,k)-datin(j,k);endendendr.=1也i丨maxdjj,ije1,2.；dijVqdij當(dāng)時d稱海明(Hamming)距離;當(dāng)q=2時，d稱歐氏(Euclfd)距離g+jfunctionR=coesimhm(datin)%聚類分析相關(guān)系數(shù)計算公式Hamming距離法%R=coesimhm(datin)error(nargchk(1,1,nargin)NM=size(datin);R=zeros(N,N);max=O;fori

9、=1:Nforj=1:NR(i,j)=sum(abs(datin(i,:)-datin(j,:);ifmaxvR(i,j)max=R(i,j);endendendfori=1:Nforj=1:NR(i,j)=1-R(i,j)/max;endendfunctionR=coesimeu(datin)%聚類分析相關(guān)系數(shù)計算公式一一Euclid歐氏距離法%R=coesimeu(datin)error(nargchk(1,1,nargin)NM=size(datin);R=zeros(N,N);max=0;fori=1:Nforj=1:NR(i,j)=sqrt(sum(datin(i,:)-datin(

10、j,:)42);ifmaxvR(i,j)max=R(i,j);endendendfori=1:Nforj=1:NR(i,j)=1-R(i,j)/max;endendfunctionR=coesimntv(datin)%聚類分析相關(guān)系數(shù)計算公式一一NTV法%R=coesimntv(datin)error(nargchk(1,1,nargin)NM=size(datin);R=zeros(N,N);fori=1:Nforj=1:NR(i,j)=1-sum(abs(datin(i,:)-datin(j,:)/sum(max(datin(i,:);datin(j,:);endendfunctionRP

11、=rpower(datin,P)%模糊關(guān)系乘法%RP=rpower(R,P)%R為模糊關(guān)系，P為R自乘的次數(shù)error(nargchk(2,2,nargin)if(P=floor(P),error(Pmustbeaninteger.);endNM=size(datin);if(N=M),error(Matrixlengthofeachdemensionmustagree.);endRP=datin;whileP1fori=1:Nforj=1:NRP(i,j)=max(min(RP(i,:);datin(:,j);endendP=P-1;endfunctionRKK=fuzequmat(R)%Fuzzyequivalentmatrix求模糊等價矩