package com.intel.daal.examples.svm;
import java.nio.FloatBuffer;
import com.intel.daal.algorithms.classifier.training.InputId;
import com.intel.daal.algorithms.classifier.prediction.PredictionResult;
import com.intel.daal.algorithms.classifier.prediction.PredictionResultId;
import com.intel.daal.algorithms.classifier.training.TrainingResultId;
import com.intel.daal.algorithms.classifier.prediction.NumericTableInputId;
import com.intel.daal.algorithms.classifier.prediction.ModelInputId;
import com.intel.daal.algorithms.multi_class_classifier.training.TrainingMethod;
import com.intel.daal.algorithms.multi_class_classifier.prediction.*;
import com.intel.daal.algorithms.multi_class_classifier.Model;
import com.intel.daal.algorithms.multi_class_classifier.ModelBuilder;
import com.intel.daal.data_management.data.NumericTable;
import com.intel.daal.data_management.data.HomogenNumericTable;
import com.intel.daal.data_management.data.MergedNumericTable;
import com.intel.daal.data_management.data_source.DataSource;
import com.intel.daal.data_management.data_source.FileDataSource;
import com.intel.daal.examples.utils.Service;
import com.intel.daal.services.DaalContext;
class SVMMultiClassModelBuilder {
private static final String [] trainedModelsFileNames = { "../data/batch/svm_multi_class_trained_model_01.csv",
"../data/batch/svm_multi_class_trained_model_02.csv",
"../data/batch/svm_multi_class_trained_model_12.csv" };
private static final String testDatasetFileName = "../data/batch/multiclass_iris_train.csv";
private static final long nFeatures = 4;
private static final long nClasses = 3;
private static final float [] biases = {-0.774F, -1.507F, -7.559F};
static Model model;
static NumericTable results;
static NumericTable testGroundTruth;
private static DaalContext context = new DaalContext();
public static void main(String[] args) throws java.io.FileNotFoundException, java.io.IOException {
buildModelFromTraining();
testModel();
printResults();
context.dispose();
}
public static void buildModelFromTraining() {
ModelBuilder multiBuilder = new ModelBuilder(context, TrainingMethod.oneAgainstOne, nFeatures, nClasses);
long imodel = 0;
for (long iClass = 1; iClass < nClasses; iClass++) {
for (long jClass = 0; jClass < iClass; jClass++, imodel++) {
FileDataSource trainDataSource = new FileDataSource(context, trainedModelsFileNames[(int)imodel],
DataSource.DictionaryCreationFlag.DoDictionaryFromContext,
DataSource.NumericTableAllocationFlag.NotAllocateNumericTable);
NumericTable supportVectors = new HomogenNumericTable(context, Float.class, nFeatures, 0, NumericTable.AllocationFlag.DoNotAllocate);
NumericTable classificationCoefficients = new HomogenNumericTable(context, Float.class, 1, 0, NumericTable.AllocationFlag.DoNotAllocate);
MergedNumericTable mergedData = new MergedNumericTable(context);
mergedData.addNumericTable(supportVectors);
mergedData.addNumericTable(classificationCoefficients);
trainDataSource.loadDataBlock(mergedData);
long nSV = supportVectors.getNumberOfRows();
com.intel.daal.algorithms.svm.ModelBuilder modelBuilder =
new com.intel.daal.algorithms.svm.ModelBuilder(context, Float.class, nFeatures, nSV);
FloatBuffer bufferSupportVectors = FloatBuffer.allocate(0);
bufferSupportVectors = supportVectors.getBlockOfRows(0, nSV, bufferSupportVectors);
modelBuilder.setSupportVectors(bufferSupportVectors, nSV*nFeatures);
FloatBuffer bufferClassCoef = FloatBuffer.allocate(0);
bufferClassCoef = classificationCoefficients.getBlockOfRows(0, nSV, bufferClassCoef);
modelBuilder.setClassificationCoefficients(bufferClassCoef, nSV);
modelBuilder.setBias(biases[(int)imodel]);
multiBuilder.setTwoClassClassifierModel(jClass, iClass, modelBuilder.getModel());
}
}
model = multiBuilder.getModel();
}
private static void testModel() {
FileDataSource testDataSource = new FileDataSource(context, testDatasetFileName,
DataSource.DictionaryCreationFlag.DoDictionaryFromContext,
DataSource.NumericTableAllocationFlag.NotAllocateNumericTable);
NumericTable testData = new HomogenNumericTable(context, Float.class, nFeatures, 0, NumericTable.AllocationFlag.DoNotAllocate);
testGroundTruth = new HomogenNumericTable(context, Float.class, 1, 0, NumericTable.AllocationFlag.DoNotAllocate);
MergedNumericTable mergedData = new MergedNumericTable(context);
mergedData.addNumericTable(testData);
mergedData.addNumericTable(testGroundTruth);
testDataSource.loadDataBlock(mergedData);
PredictionBatch multiPredict = new PredictionBatch(context, Float.class, PredictionMethod.multiClassClassifierWu, nClasses);
com.intel.daal.algorithms.svm.prediction.PredictionBatch twoClassPrediction = new com.intel.daal.algorithms.svm.prediction.PredictionBatch(
context, Float.class, com.intel.daal.algorithms.svm.prediction.PredictionMethod.defaultDense);
multiPredict.parameter.setPrediction(twoClassPrediction);
multiPredict.input.set(NumericTableInputId.data, testData);
multiPredict.input.set(ModelInputId.model, model);
PredictionResult predictionResult = multiPredict.compute();
results = predictionResult.get(PredictionResultId.prediction);
}
private static void printResults() {
Service.printClassificationResult(testGroundTruth, results, "Ground truth", "Classification results",
"Multi-class SVM classification sample program results (first 20 observations):", 20);
System.out.println("");
}
}