matlab.rar_Fisher线性判别_二次多项式_二次聚类_多项式回归_多项式系统

function CollectAnaly(X) format long; sz = size(X); N = sz(1); %样本个数 n = sz(2); D = zeros(n,n); totalClass = N; RecordClass = zeros(N,N+1); RecordClass(:,1) = ones(N,1); RecordClass(:,2) = 1:N; disp('聚类前的N个类：'); disp(RecordClass); while totalClass > 1 minClaDist = inf; for i=1:totalClass for j=i+1:totalClass distClass = DistClass(X,RecordClass,i,j,RecordClass(i,1),RecordClass(j,1)); if distClass < minClaDist minClaDist = distClass; i1 = i; j1 = j; end end end t1 = RecordClass(i1,1); t2 = RecordClass(j1,1); RecordClass(i1,(t1+2):(t1+t2+1)) = RecordClass(j1,2:(t2+1)); RecordClass(i1,1) = RecordClass(i1,1) + RecordClass(j1,1); RecordClass(j1:(totalClass-1),:) = RecordClass((j1+1):totalClass,:); RecordClass(totalClass:N,:) = zeros(N-totalClass+1,N+1); totalClass = totalClass - 1; str1 = strcat('第',num2str(N - totalClass)); str1 = strcat(str1, '次聚类，'); str1 = strcat(str1, '第'); str1 = strcat(str1, num2str(i1)); str1 = strcat(str1, '类和第'); str1 = strcat(str1, num2str(j1)); str1 = strcat(str1, '类合并：'); disp(str1); disp(RecordClass(1:totalClass,:)) end function d = DistSamp(X1,X2) format long; d = sqrt(dot( X1-X2,X1-X2)); format short; function d = DistClass(X,C1,l1,l2,n,m) format long; d = inf; for i=1:n for j=1:m dc = DistSamp(X(C1(l1,i+1),:),X(C1(l2,j+1),:)); if dc < d d = dc; end end end format short;