makewinnertable.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 [ActUnits, ActWeights] = MakeWinnerTable(Radius,MLKP);

MaxWinner=MLKP.Nrow*MLKP.Ncol;
MaxRadius=min((2*Radius+1)^2,MaxWinner);
ActUnits=zeros(MaxWinner,MaxRadius+1);
ActWeights=zeros(MaxWinner,MaxRadius);

if Radius > 0
    for Winner=1:MaxWinner
        [Neighbours,Weights] = GetNeighbours(Radius,Winner,MLKP);
        Len=length(Neighbours);
        ActUnits(Winner,1:Len+1)=[Len Neighbours];
        ActWeights(Winner,1:Len)=Weights;
    end
else
    for Winner=1:MaxWinner
        ActUnits(Winner,1:2)=[1 Winner];
        ActWeights(Winner,1)=1;
    end
end