public void SVM_kernel_poly_degree_1_smo_all_training_samples()
{
initData_dataset_linear_subgd_jason_example();
BuildSVMKernelSMO build = new BuildSVMKernelSMO();
setPrivateVariablesInBuildObject(build);
//Set params
build.SetParameters(2, .0001, .001);
build.setKernel(BuildSVMKernelSMO.Kernels.Polynomial, 1);
ModelSVMKernelSMO model =
(ModelSVMKernelSMO)build.BuildModel(
_trainingData, _attributeHeaders,
_indexTargetAttribute);
int count = 0;
for (int row = 0; row < _trainingData[0].Length; row++)
{
double[] data = GetSingleTrainingRowDataForTest(row);
double value = model.RunModelForSingleData(data);
if (value == _trainingData[_indexTargetAttribute][row])
{
count++;
}
}
//Random function is adding some randomness
Assert.IsTrue(count >= 8 && count <= 9);
}
public void SVM_kernel_smo_single_training_sample()
{
initData_dataset_linear_subgd_jason_example();
BuildSVMKernelSMO build = new BuildSVMKernelSMO();
setPrivateVariablesInBuildObject(build);
//Set params
//build.setParameters(1,1,.45);
ModelSVMKernelSMO model =
(ModelSVMKernelSMO)build.BuildModel(
_trainingData, _attributeHeaders,
_indexTargetAttribute);
double[] data = GetSingleTrainingRowDataForTest(0);
double value = model.RunModelForSingleData(data);
Assert.AreEqual(value,
_trainingData[_indexTargetAttribute][0]);
}
