leastsquaresclassifier.m.svn-base

来自「a function inside machine learning」· SVN-BASE 代码 · 共 18 行

function [trainInfo, testInfo, classifierInfo] = leastSquaresClassifier(trainX, trainY, testX, params); 
%A function to compute least squares classification on the data on binary labels. No bias term here.   

if nargin ~= 4
    fprintf('%s\n', help('leastSquaresClassifier'));
    error('Incorrect number of inputs - see above usage instructions.'); 
end

if ~binaryLabels(trainY) 
    error('Using leastSquaresClassifier with non binary labels'); 
end

[trainInfo, testInfo, classifierInfo] = leastSquaresRegressor(trainX, trainY, testX, params); 

trainInfo.predictedY = sign(trainInfo.predictedY); 
testInfo.predictedY = sign(testInfo.predictedY);