SG++
sgpp::datadriven::LearnerPiecewiseConstantSmoothedRegression Class Reference

#include <LearnerPiecewiseConstantSmoothedRegression.hpp>

Public Member Functions

sgpp::base::OperationMatrixcomputeRegularizationMatrix (sgpp::base::Grid &grid)
 generates the regularization matrix More...
 
 LearnerPiecewiseConstantSmoothedRegression (sgpp::base::RegularGridConfiguration &gridConfig, sgpp::base::AdpativityConfiguration &adaptivityConfig, sgpp::solver::SLESolverConfiguration &solverConfig, sgpp::datadriven::RegularizationConfiguration &regularizationConfig, bool verbose)
 Constructor. More...
 
double optimizeLambdaCV (size_t kFold)
 Does cross-validation to obtain a suitable regularization parameter. More...
 
void splitset (std::vector< sgpp::base::DataMatrix * > &strain, std::vector< sgpp::base::DataMatrix * > &stest)
 splits the complete sample set in a set of smaller training and test samples for cross-validation. More...
 
void train (sgpp::datadriven::PiecewiseConstantRegression::Node &piecewiseRegressor, sgpp::base::Grid &grid, sgpp::base::DataVector &alpha, double lambda)
 Does the learning step on a given grid, training set and regularization parameter lambda. More...
 

Constructor & Destructor Documentation

sgpp::datadriven::LearnerPiecewiseConstantSmoothedRegression::LearnerPiecewiseConstantSmoothedRegression ( sgpp::base::RegularGridConfiguration gridConfig,
sgpp::base::AdpativityConfiguration adaptivityConfig,
sgpp::solver::SLESolverConfiguration solverConfig,
sgpp::datadriven::RegularizationConfiguration regularizationConfig,
bool  verbose 
)

Constructor.

Parameters
gridConfiggrid configuration
adaptivityConfigadaptive refinement configuration
solverConfigsolver configuration (CG)
regularizationConfigconfig for regularization operator
verbosereport additional information on the console

Member Function Documentation

OperationMatrix * sgpp::datadriven::LearnerPiecewiseConstantSmoothedRegression::computeRegularizationMatrix ( sgpp::base::Grid grid)
double sgpp::datadriven::LearnerPiecewiseConstantSmoothedRegression::optimizeLambdaCV ( size_t  kFold)

Does cross-validation to obtain a suitable regularization parameter.

void sgpp::datadriven::LearnerPiecewiseConstantSmoothedRegression::splitset ( std::vector< sgpp::base::DataMatrix * > &  strain,
std::vector< sgpp::base::DataMatrix * > &  stest 
)

splits the complete sample set in a set of smaller training and test samples for cross-validation.

Parameters
strainvector containing the training samples for cv
stestvector containing the test samples for cv

The documentation for this class was generated from the following files: