leastsquarestest.m.svn-base

a function inside machine learning

%Test least squares 
clear all; 
numExamples = 150; 
numFeatures = 10; 

noise = 0.5; 
tol = 10^-5; 

X = rand(numExamples, numFeatures); 
X = centerData(X); 
X = normalise(X); 
q = rand(numFeatures, 1); 
Y = centerData(X*q + noise*(rand(numExamples, 1)-0.5));


d = data; 
d = addDataField(d, 'X', X, 'examples'); 
d = addDataField(d, 'Y', Y, 'labels'); 

[trainData, testData] = splitData2(d, 2/3); 

params = struct; 

[classifierInfo, trainInfo] = leastSquaresRegressorTrain(trainData, params);
[testInfo, predictionInfo] = leastSquaresRegressorPredict(trainData, testData, classifierInfo, params);

correlation(getDataFieldValue(trainData, 'Y'), trainInfo.predictedY)
correlation(getDataFieldValue(testData, 'Y'), testInfo.predictedY)

[trainInfo, testInfo, classifierInfo] = leastSquaresRegressor(getDataFieldValue(trainData, 'X'), getDataFieldValue(trainData, 'Y'), getDataFieldValue(testData, 'X'), params); 

correlation(getDataFieldValue(trainData, 'Y'), trainInfo.predictedY)
correlation(getDataFieldValue(testData, 'Y'), testInfo.predictedY)

[testInfo, predictionInfo] = leastSquaresRegressorPredict(trainData, trainData, classifierInfo, params);
correlation(getDataFieldValue(trainData, 'Y'), testInfo.predictedY)

clearAllFields(d); 

%Test multi label case 
numLabels = 5; 
q = rand(numFeatures, numLabels); 
Y = centerData(X*q + noise*(rand(numExamples, numLabels)-0.5));

d = data; 
d = addDataField(d, 'X', X, 'examples'); 
d = addDataField(d, 'Y', Y, 'labels'); 

[trainData, testData] = splitData2(d, 2/3); 

[classifierInfo, trainInfo] = leastSquaresRegressorTrain(trainData, params);
[testInfo, predictionInfo] = leastSquaresRegressorPredict(trainData, testData, classifierInfo, params);

correlation(getDataFieldValue(trainData, 'Y'), trainInfo.predictedY)
correlation(getDataFieldValue(testData, 'Y'), testInfo.predictedY)