/// <summary>Sets multi-class Support Vector Machine classifier</summary>
/// <remarks><para>Removes any existing classifier and sets mSVM with specified settings</para></remarks>
/// <param name="loss">Loss function to set</param>
/// <param name="model">Model to be used with SVM classifier</param>
/// <seealso cref="aceOperationSetExecutorBase"/>
public void aceOperation_setmSVM(
[Description("Loss function to set")] Loss loss = Loss.L2,
[Description("Model to be used with SVM classifier")] mSVMModels model = mSVMModels.linear)
{
data.classifierSettings.type = imbNLP.Toolkit.Classifiers.ClassifierType.multiClassSVM;
data.classifierSettings.lossFunctionForTraining = loss;
data.classifierSettings.svmModel = model;
}
public override void Deploy(ClassifierSettings _setup)
{
setup = _setup;
model = setup.svmModel;
if (model == mSVMModels.linear)
{
// Create a one-vs-one multi-class SVM learning algorithm
teacher = new MulticlassSupportVectorLearning <Linear>()
{
// using LIBLINEAR's L2-loss SVC dual for each SVM
Learner = (p) => new LinearDualCoordinateDescent()
{
Loss = setup.lossFunctionForTraining
}
};
}
if (model == mSVMModels.gaussian)
{
// Create the multi-class learning algorithm for the machine
teacherGaussian = new MulticlassSupportVectorLearning <Gaussian>()
{
// Configure the learning algorithm to use SMO to train the
// underlying SVMs in each of the binary class subproblems.
Learner = (param) => new SequentialMinimalOptimization <Gaussian>()
{
// Estimate a suitable guess for the Gaussian kernel's parameters.
// This estimate can serve as a starting point for a grid search.
UseKernelEstimation = true
}
};
}
//teacher.UseKernelEstimation = true;
// The following line is only needed to ensure reproducible results. Please remove it to enable full parallelization
// teacher.ParallelOptions.MaxDegreeOfParallelism = 1; // (Remove, comment, or change this line to enable full parallelism)
}