meanperclass.m

有监督自组织映射-偏最小二乘算法(A supervised self-organising map–partial least squares algorithm),可以用语多变量数据的回归分析

function [MeanVal] = MeanPerClass(XMap,YMap);

[NvarX,Nunit]=size(XMap);
[NvarY,Nunit]=size(YMap);
MeanVal=zeros(NvarY,NvarX);
NormFact=zeros(1,NvarY);

for iunit=1:Nunit
    [Value,Pos]=max(YMap(:,iunit));
    if Value > 0.95
        MeanVal(Pos,:)=MeanVal(Pos,:)+XMap(:,iunit)';
        NormFact(Pos)=NormFact(Pos)+1;
    end
end

for i=1:NvarY
    MeanVal(i,:)=MeanVal(i,:)/NormFact(i);
end