svmcTrain

Training SVM (support vector machine) classifier

Contents

Syntax

Description

cPrm=svmcTrain(DS) returns the parameters of SVM (support vector machine) based on the given dataset DS.

opt=svmcTrain('defaultOpt') returns the default options for SVM. You can modify the options if necessary, and send the options for training a SVM:

cPrm=svmcTrain(DS, opt)

Note that this function calls the mex files of libsvm which are available at "http://www.csie.ntu.edu.tw/~cjlin/libsvm/".

Example

DS=prData('iris');
trainSet.input=DS.input(:, 1:2:end); trainSet.output=DS.output(:, 1:2:end);
testSet.input=DS.input(:, 2:2:end);  testSet.output=DS.output(:, 2:2:end);
[cPrm, logLike1, recogRate1]=svmcTrain(trainSet);
[computedClass, logLike2, recogRate2, hitIndex]=svmcEval(testSet, cPrm);
fprintf('Inside recog. rate = %g%%\n', recogRate1*100);
fprintf('Outside recog. rate = %g%%\n', recogRate2*100);

Inside recog. rate = 100%
Outside recog. rate = 90.6667%

See Also

svmcEval.

Published with MATLAB® R2018a