plspred.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 [YPredict] = PlsPred(X, Y, B, MLKP);
%
% Predicts output based on a known model (B coefficients)
%
% Format: [Ypredict] = plspred(X, Y, B, MLKP)
%
% X:    input data
% Y:    output data
% B:    matrix containing the regression coefficients
% MLKP: parameter structure
%
% Ypredict: prediction of the output date based in X and B
%
[Nobj,NvarX] = size(X);
[Nobj,NvarY] = size(Y);
if MLKP.PLSMode == 'SIMPLS'
    XTrainMC = (X-MLKP.PLSMX(ones(Nobj,1),:));
    YPredictMC = XTrainMC * B;
    YPredict = (YPredictMC+MLKP.PLSMY(ones(Nobj,1),:));
else
    YPredict = X * B;
end