sgpp::datadriven::LearnerSGDE Class Reference

#include <LearnerSGDE.hpp>

Inheritance diagram for sgpp::datadriven::LearnerSGDE:

Public Member Functions
void	cov (base::DataMatrix &cov, base::DataMatrix *bounds=nullptr) override
	WARNING: Not yet implemented. More...
virtual double	getAccuracy (base::DataMatrix &testDataset, const base::DataVector &referenceLabels, const double threshold)
	Computes the classification accuracy. More...
virtual double	getAccuracy (const base::DataVector &referenceLabels, const double threshold, const base::DataVector &predictedLabels)
	Computes the classification accuracy. More...
size_t	getDim () override
	get number of dimensions More...
virtual double	getError (base::DataMatrix &data, const base::DataVector &labels, const double threshold, std::string errorType)
	Error evaluation required for convergence-based refinement. More...
virtual base::Grid *	getGrid ()
	returns the grid More...
size_t	getNsamples () override
	get number of samples More...
std::shared_ptr< base::DataVector >	getSamples (size_t dim) override
	returns the samples in the given dimension More...
std::shared_ptr< base::DataMatrix >	getSamples () override
	returns the complete sample set More...
std::shared_ptr< base::Grid >	getSharedGrid ()
std::shared_ptr< base::DataVector >	getSharedSurpluses ()
virtual base::DataVector *	getSurpluses ()
	returns the surpluses More...
void	initialize (base::DataMatrix &samples) override
	Create grid and perform cross-validation if enabled. More...
	LearnerSGDE (sgpp::base::RegularGridConfiguration &gridConfig, sgpp::base::AdaptivityConfiguration &adaptivityConfig, sgpp::solver::SLESolverConfiguration &solverConfig, sgpp::datadriven::RegularizationConfiguration &regularizationConfig, CrossvalidationConfiguration &crossvalidationConfig)
	Constructor. More...
	LearnerSGDE (LearnerSGDEConfiguration &learnerSGDEConfig)
	LearnerSGDE (const LearnerSGDE &learnerSGDE)
double	mean () override
	This method computes the mean of the density function. More...
double	pdf (base::DataVector &x) override
	This methods evaluates the sparse grid density at a single point. More...
void	pdf (base::DataMatrix &points, base::DataVector &res) override
	Evaluation of the sparse grid density at a set of points. More...
virtual void	predict (base::DataMatrix &testDataset, base::DataVector &predictedLabels)
	Predicts class labels based on the trained model. More...
virtual void	storeResults (base::DataMatrix &testDataset)
	Stores classified data, grids and density function evaluations to csv files. More...
virtual void	train (base::Grid &grid, base::DataVector &alpha, base::DataMatrix &trainData, double lambdaReg)
	Does the learning step (i.e. More...
virtual void	train ()
	Learns the data. More...
virtual void	trainOnline (base::DataVector &labels, base::DataMatrix &testData, base::DataVector &testLabels, base::DataMatrix *validData, base::DataVector *validLabels, base::DataVector &classLabels, size_t maxDataPasses, std::string refType, std::string refMonitor, size_t refPeriod, double accDeclineThreshold, size_t accDeclineBufferSize, size_t minRefInterval, bool usePrior)
	Performs the sparse grid density estimation via online learning. More...
double	variance () override
	Computes the variance of the density function. More...
virtual	~LearnerSGDE ()
Public Member Functions inherited from sgpp::datadriven::DensityEstimator
virtual void	corrcoef (base::DataMatrix &corr, base::DataMatrix *bounds=nullptr)
double	crossEntropy (sgpp::base::DataMatrix &samples)
	DensityEstimator ()
virtual double	std_deviation ()
virtual	~DensityEstimator ()

Public Attributes
base::DataVector	avgErrors
double	error

Protected Member Functions
base::OperationMatrix *	computeRegularizationMatrix (base::Grid &grid)
	generates the regularization matrix More...
double	computeResidual (base::Grid &grid, base::DataVector &alpha, base::DataMatrix &test, double lambdaReg)
	Compute the residual for a given test data set on a learned grid. More...
std::shared_ptr< base::Grid >	createRegularGrid ()
	generates a regular grid More...
double	mean (base::Grid &grid, base::DataVector &alpha)
double	optimizeLambdaCV ()
	Does cross-validation to obtain a suitable regularization parameter. More...
void	splitset (std::vector< std::shared_ptr< base::DataMatrix >> &strain, std::vector< std::shared_ptr< base::DataMatrix >> &stest)
	splits the complete sample set in a set of smaller training and test samples for cross-validation. More...
double	variance (base::Grid &grid, base::DataVector &alpha)

Protected Attributes
sgpp::base::AdaptivityConfiguration	adaptivityConfig
std::shared_ptr< base::DataVector >	alpha
std::map< int, std::shared_ptr< base::DataVector > >	alphas
std::map< int, size_t >	appearances
sgpp::datadriven::CrossvalidationConfiguration	crossvalidationConfig
std::shared_ptr< base::Grid >	grid
sgpp::base::RegularGridConfiguration	gridConfig
std::map< int, std::shared_ptr< base::Grid > >	grids
double	lambdaReg
std::map< int, double >	priors
sgpp::datadriven::RegularizationConfiguration	regularizationConfig
sgpp::solver::SLESolverConfiguration	solverConfig
std::shared_ptr< base::DataMatrix >	trainData
std::shared_ptr< base::DataVector >	trainLabels
bool	usePrior

Constructor & Destructor Documentation

LearnerSGDE() [1/3]

sgpp::datadriven::LearnerSGDE::LearnerSGDE	(	sgpp::base::RegularGridConfiguration &	gridConfig,
		sgpp::base::AdaptivityConfiguration &	adaptivityConfig,
		sgpp::solver::SLESolverConfiguration &	solverConfig,
		sgpp::datadriven::RegularizationConfiguration &	regularizationConfig,
		CrossvalidationConfiguration &	crossvalidationConfig
	)

Constructor.

Parameters

gridConfig	grid configuration
adaptivityConfig	adaptive refinement configuration
solverConfig	solver configuration (CG)
regularizationConfig	config for regularization operator
crossvalidationConfig	configuration for the cross validation

LearnerSGDE() [2/3]

sgpp::datadriven::LearnerSGDE::LearnerSGDE ( LearnerSGDEConfiguration &  learnerSGDEConfig )

explicit

LearnerSGDE() [3/3]

sgpp::datadriven::LearnerSGDE::LearnerSGDE

(

const LearnerSGDE &

learnerSGDE

)

References adaptivityConfig, alpha, crossvalidationConfig, error, grid, gridConfig, lambdaReg, regularizationConfig, solverConfig, trainData, and usePrior.

~LearnerSGDE()

sgpp::datadriven::LearnerSGDE::~LearnerSGDE ( )

virtual

Member Function Documentation

computeRegularizationMatrix()

base::OperationMatrix * sgpp::datadriven::LearnerSGDE::computeRegularizationMatrix ( base::Grid &  grid )

protected

generates the regularization matrix

Parameters

grid

grid

References python.utils.pca_normalize_dataset::C, sgpp::op_factory::createOperationIdentity(), sgpp::op_factory::createOperationLaplace(), sgpp::datadriven::Identity, sgpp::datadriven::Laplace, regularizationConfig, and sgpp::datadriven::RegularizationConfiguration::type_.

Referenced by computeResidual(), train(), and trainOnline().

computeResidual()

double sgpp::datadriven::LearnerSGDE::computeResidual ( base::Grid &  grid, base::DataVector &  alpha, base::DataMatrix &  test, double  lambdaReg  )

protected

Compute the residual for a given test data set on a learned grid.

$|(A - lambda C) alpha - 1/n B|$

This is used as quality criterion for the estimated density.

Parameters

grid	grid
alpha	coefficient vector
test	test set
lambdaReg	regularization parameters

Returns

References python.utils.pca_normalize_dataset::C, computeRegularizationMatrix(), sgpp::base::Grid::getSize(), and python.statsfileInfo::i.

Referenced by optimizeLambdaCV().

cov()

void sgpp::datadriven::LearnerSGDE::cov ( base::DataMatrix &  cov, base::DataMatrix *  bounds = nullptr  )

overridevirtual

WARNING: Not yet implemented.

Implements sgpp::datadriven::DensityEstimator.

References alpha, sgpp::op_factory::createOperationCovariance(), and grid.

createRegularGrid()

std::shared_ptr< base::Grid > sgpp::datadriven::LearnerSGDE::createRegularGrid ( )

protected

generates a regular grid

Returns

grid

References sgpp::base::Grid::createLinearBoundaryGrid(), sgpp::base::Grid::createLinearGrid(), sgpp::base::GeneralGridConfiguration::dim_, sgpp::base::GeneralGridConfiguration::filename_, gridConfig, sgpp::base::GeneralGridConfiguration::level_, sgpp::base::Linear, sgpp::base::LinearBoundary, sgpp::base::LinearL0Boundary, sgpp::base::GeneralGridConfiguration::type_, and sgpp::base::Grid::unserialize().

Referenced by initialize(), and optimizeLambdaCV().

getAccuracy() [1/2]

double sgpp::datadriven::LearnerSGDE::getAccuracy ( base::DataMatrix &  testDataset, const base::DataVector &  referenceLabels, const double  threshold  )

virtual

Computes the classification accuracy.

Parameters

testDataset	The data for which class labels should be predicted
referenceLabels	The corresponding actual class labels
threshold	The decision threshold (e.g. for class labels -1, 1 -> threshold = 0)

Returns

The resulting accuracy

References sgpp::base::DataMatrix::getNrows(), and predict().

Referenced by getError(), and trainOnline().

getAccuracy() [2/2]

double sgpp::datadriven::LearnerSGDE::getAccuracy ( const base::DataVector &  referenceLabels, const double  threshold, const base::DataVector &  predictedLabels  )

virtual

Computes the classification accuracy.

Parameters

referenceLabels	The actual class labels
threshold	The decision threshold (e.g. for class labels -1, 1 -> threshold = 0)
predictedLabels	The predicted class labels

Returns

The resulting accuracy

References sgpp::base::DataVector::get(), sgpp::base::DataVector::getSize(), and python.statsfileInfo::i.

getDim()

size_t sgpp::datadriven::LearnerSGDE::getDim ( )

overridevirtual

get number of dimensions

Implements sgpp::datadriven::DensityEstimator.

References sgpp::base::GeneralGridConfiguration::dim_, and gridConfig.

Referenced by python.uq.dists.SGDEdist.SGDEdist::__str__(), and python.uq.dists.KDEDist.KDEDist::getBandwidths().

getError()

double sgpp::datadriven::LearnerSGDE::getError ( base::DataMatrix &  data, const base::DataVector &  labels, const double  threshold, std::string  errorType  )

virtual

Error evaluation required for convergence-based refinement.

Parameters

data	The data points to measure the error on
labels	The corresponding class labels
threshold	The decision threshold (e.g. for class labels -1, 1 -> threshold = 0)
errorType	The error type (only "Acc" possible, i.e. classification error based on accuracy)

Returns

The error evaluation

References getAccuracy().

Referenced by trainOnline().

getGrid()

base::Grid * sgpp::datadriven::LearnerSGDE::getGrid ( )

virtual

returns the grid

References grid.

getNsamples()

size_t sgpp::datadriven::LearnerSGDE::getNsamples ( )

overridevirtual

get number of samples

Implements sgpp::datadriven::DensityEstimator.

References trainData.

Referenced by getSamples(), and optimizeLambdaCV().

getSamples() [1/2]

std::shared_ptr< base::DataVector > sgpp::datadriven::LearnerSGDE::getSamples ( size_t  dim )

overridevirtual

returns the samples in the given dimension

Parameters

dim

Implements sgpp::datadriven::DensityEstimator.

References getNsamples(), and trainData.

getSamples() [2/2]

std::shared_ptr< base::DataMatrix > sgpp::datadriven::LearnerSGDE::getSamples ( )

overridevirtual

returns the complete sample set

Implements sgpp::datadriven::DensityEstimator.

References trainData.

getSharedGrid()

std::shared_ptr< base::Grid > sgpp::datadriven::LearnerSGDE::getSharedGrid

(

)

References grid.

Referenced by sgpp::datadriven::RosenblattTransformation::initialize().

getSharedSurpluses()

std::shared_ptr< base::DataVector > sgpp::datadriven::LearnerSGDE::getSharedSurpluses

(

)

References alpha.

Referenced by sgpp::datadriven::RosenblattTransformation::initialize().

getSurpluses()

base::DataVector * sgpp::datadriven::LearnerSGDE::getSurpluses ( )

virtual

returns the surpluses

References alpha.

initialize()

void sgpp::datadriven::LearnerSGDE::initialize ( base::DataMatrix &  samples )

overridevirtual

Create grid and perform cross-validation if enabled.

Parameters

samples

DataMatrix (nrows = number of samples, ncols = dimensionality)

Implements sgpp::datadriven::DensityEstimator.

References alpha, createRegularGrid(), crossvalidationConfig, sgpp::base::GeneralGridConfiguration::dim_, sgpp::datadriven::CrossvalidationConfiguration::enable_, sgpp::base::DataMatrix::getNcols(), grid, gridConfig, sgpp::datadriven::CrossvalidationConfiguration::lambda_, lambdaReg, optimizeLambdaCV(), parabolasimple::samples, and trainData.

Referenced by sgpp::datadriven::RosenblattTransformation::initialize().

mean() [1/2]

double sgpp::datadriven::LearnerSGDE::mean ( )

overridevirtual

This method computes the mean of the density function.

Implements sgpp::datadriven::DensityEstimator.

References alpha, and grid.

Referenced by python.uq.analysis.mc.MCAnalysis.MCAnalysis::computeMoments(), python.uq.analysis.asgc.ASGCAnalysis.ASGCAnalysis::computeMoments(), and variance().

mean() [2/2]

double sgpp::datadriven::LearnerSGDE::mean ( base::Grid &  grid, base::DataVector &  alpha  )

protected

References sgpp::op_factory::createOperationFirstMoment(), and sgpp::base::OperationFirstMoment::doQuadrature().

Referenced by python.uq.analysis.mc.MCAnalysis.MCAnalysis::computeMoments(), and python.uq.analysis.asgc.ASGCAnalysis.ASGCAnalysis::computeMoments().

optimizeLambdaCV()

double sgpp::datadriven::LearnerSGDE::optimizeLambdaCV ( )

protected

Does cross-validation to obtain a suitable regularization parameter.

References alpha, computeResidual(), createRegularGrid(), crossvalidationConfig, getNsamples(), grid, python.statsfileInfo::i, python.utils.statsfile2gnuplot::j, sgpp::datadriven::CrossvalidationConfiguration::kfold_, sgpp::datadriven::CrossvalidationConfiguration::lambdaEnd_, sgpp::datadriven::CrossvalidationConfiguration::lambdaStart_, sgpp::datadriven::CrossvalidationConfiguration::lambdaSteps_, sgpp::datadriven::CrossvalidationConfiguration::logScale_, sgpp::base::DataVector::max(), sgpp::base::DataVector::min(), sgpp::base::DataVector::resizeZero(), sgpp::datadriven::CrossvalidationConfiguration::silent_, splitset(), and train().

Referenced by initialize().

pdf() [1/2]

double sgpp::datadriven::LearnerSGDE::pdf ( base::DataVector &  x )

overridevirtual

This methods evaluates the sparse grid density at a single point.

Parameters

x	DataVector length equal to dimensionality

Implements sgpp::datadriven::DensityEstimator.

References alpha, sgpp::op_factory::createOperationEval(), and sgpp::base::OperationEval::eval().

pdf() [2/2]

void sgpp::datadriven::LearnerSGDE::pdf ( base::DataMatrix &  points, base::DataVector &  res  )

overridevirtual

Evaluation of the sparse grid density at a set of points.

Parameters

points	DataMatrix (nrows = number of samples, ncols = dimensionality)
res	DataVector (size = number of samples) where the results are stored

Implements sgpp::datadriven::DensityEstimator.

References alpha, sgpp::op_factory::createOperationMultipleEval(), sgpp::base::OperationMultipleEval::eval(), and grid.

predict()

void sgpp::datadriven::LearnerSGDE::predict ( base::DataMatrix &  testDataset, base::DataVector &  predictedLabels  )

virtual

Predicts class labels based on the trained model.

Parameters

testDataset	The data for which class labels should be predicted
predictedLabels	The predicted class labels

References alphas, sgpp::op_factory::createOperationEval(), chess::dim, g, sgpp::base::DataMatrix::getNcols(), sgpp::base::DataMatrix::getNrows(), sgpp::base::DataMatrix::getRow(), grids, python.statsfileInfo::i, priors, sgpp::base::DataVector::set(), and usePrior.

Referenced by getAccuracy(), and storeResults().

splitset()

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

protected

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

Parameters

strain	vector containing the training samples for cv
stest	vector containing the test samples for cv

References crossvalidationConfig, python.statsfileInfo::i, python.utils.statsfile2gnuplot::j, sgpp::datadriven::CrossvalidationConfiguration::kfold_, friedman::p, chess::r, create_scripts::s, sgpp::datadriven::CrossvalidationConfiguration::seed_, sgpp::datadriven::CrossvalidationConfiguration::shuffle_, sgpp::datadriven::CrossvalidationConfiguration::silent_, and analyse_erg::tmp.

Referenced by optimizeLambdaCV().

storeResults()

void sgpp::datadriven::LearnerSGDE::storeResults ( base::DataMatrix &  testDataset )

virtual

Stores classified data, grids and density function evaluations to csv files.

Parameters

testDataset

The data for which class labels should be predicted

References alphas, sgpp::base::DataMatrix::appendRow(), sgpp::base::HashGridStorage::begin(), sgpp::op_factory::createOperationEval(), sgpp::base::HashGridStorage::end(), g, sgpp::base::DataVector::get(), sgpp::base::HashGridStorage::getCoordinates(), sgpp::base::DataMatrix::getNcols(), sgpp::base::DataMatrix::getNrows(), sgpp::base::DataMatrix::getRow(), sgpp::base::DataVector::getSize(), grid, grids, python.statsfileInfo::i, python.utils.statsfile2gnuplot::j, predict(), and sgpp::base::DataVector::set().

train() [1/2]

void sgpp::datadriven::LearnerSGDE::train ( base::Grid &  grid, base::DataVector &  alpha, base::DataMatrix &  trainData, double  lambdaReg  )

virtual

Does the learning step (i.e.

computes pdf) on a given grid, training set and regularization parameter lambda

Parameters

grid	grid
alpha	coefficient vector
trainData	sample set
lambdaReg	regularization parameter

References adaptivityConfig, python.utils.pca_normalize_dataset::C, computeRegularizationMatrix(), sgpp::op_factory::createOperationEval(), crossvalidationConfig, chess::dim, sgpp::solver::SLESolverConfiguration::eps_, sgpp::datadriven::DensitySystemMatrix::generateb(), sgpp::base::DataVector::get(), sgpp::base::Grid::getGenerator(), sgpp::base::DataMatrix::getNcols(), sgpp::base::HashGridStorage::getPoint(), sgpp::solver::SGSolver::getResiduum(), sgpp::base::DataVector::getSize(), sgpp::base::Grid::getSize(), sgpp::base::HashGridPoint::getStandardCoordinates(), sgpp::base::Grid::getStorage(), python.utils.sg_projections::gridStorage, python.statsfileInfo::i, sgpp::solver::SLESolverConfiguration::maxIterations_, sgpp::base::AdaptivityConfiguration::noPoints_, sgpp::base::AdaptivityConfiguration::numRefinements_, friedman::p, sgpp::base::GridGenerator::refine(), sgpp::base::DataVector::setAll(), sgpp::datadriven::CrossvalidationConfiguration::silent_, sgpp::solver::ConjugateGradients::solve(), solverConfig, sgpp::solver::SLESolverConfiguration::threshold_, and sgpp::base::AdaptivityConfiguration::threshold_.

Referenced by sgpp::datadriven::RosenblattTransformation::initialize().

train() [2/2]

void sgpp::datadriven::LearnerSGDE::train ( )

virtual

Learns the data.

References alpha, grid, lambdaReg, and trainData.

Referenced by optimizeLambdaCV().

trainOnline()

void sgpp::datadriven::LearnerSGDE::trainOnline ( base::DataVector &  labels, base::DataMatrix &  testData, base::DataVector &  testLabels, base::DataMatrix *  validData, base::DataVector *  validLabels, base::DataVector &  classLabels, size_t  maxDataPasses, std::string  refType, std::string  refMonitor, size_t  refPeriod, double  accDeclineThreshold, size_t  accDeclineBufferSize, size_t  minRefInterval, bool  usePrior  )

virtual

Performs the sparse grid density estimation via online learning.

Parameters

labels	The training labels
testData	The test data
testLabels	The corresponding test labels
validData	The validation data
validLabels	The corresponding validation labels
classLabels	The ocurring class labels (e.g. -1,1)
maxDataPasses	The number of passes over the whole training data
refType	The refinement indicator (surplus, zero-crossings or data-based)
refMonitor	The refinement strategy (periodic or convergence-based)
refPeriod	The refinement interval (if periodic refinement is chosen)
accDeclineThreshold	The convergence threshold (if convergence-based refinement is chosen)
accDeclineBufferSize	The number of accuracy measurements which are used to check convergence (if convergence-based refinement is chosen)
minRefInterval	The minimum number of data points (or data batches) which have to be processed before next refinement can be scheduled (if convergence-based refinement is chosen)
usePrior	Specifies whether prior probabilities should be used to predict class labels

References adaptivityConfig, alpha, alphas, appearances, sgpp::base::DataVector::append(), avgErrors, python.utils.pca_normalize_dataset::C, computeRegularizationMatrix(), sgpp::base::Grid::createLinearGrid(), sgpp::base::Grid::createModLinearGrid(), sgpp::op_factory::createOperationEval(), chess::dim, sgpp::base::GeneralGridConfiguration::dim_, sgpp::solver::SLESolverConfiguration::eps_, error, g, sgpp::datadriven::DensitySystemMatrix::generateb(), getAccuracy(), getError(), sgpp::base::HashGridStorage::getPoint(), sgpp::base::DataVector::getSize(), sgpp::base::HashGridPoint::getStandardCoordinates(), grid, gridConfig, grids, python.utils.sg_projections::gridStorage, python.statsfileInfo::i, python.utils.statsfile2gnuplot::j, lambdaReg, sgpp::base::GeneralGridConfiguration::level_, sgpp::base::Linear, sgpp::solver::SLESolverConfiguration::maxIterations_, sgpp::base::ModLinear, sgpp::base::AdaptivityConfiguration::noPoints_, sgpp::base::AdaptivityConfiguration::numRefinements_, friedman::p, sgpp::datadriven::MultiGridRefinementFunctor::preComputeEvaluations(), priors, sgpp::datadriven::RefinementMonitor::pushToBuffer(), sgpp::datadriven::RefinementMonitor::refinementsNecessary(), sgpp::datadriven::MultiGridRefinementFunctor::setGridIndex(), sgpp::solver::ConjugateGradients::solve(), solverConfig, sgpp::solver::SLESolverConfiguration::threshold_, sgpp::base::AdaptivityConfiguration::threshold_, trainData, trainLabels, sgpp::base::GeneralGridConfiguration::type_, and usePrior.

variance() [1/2]

double sgpp::datadriven::LearnerSGDE::variance ( )

overridevirtual

Computes the variance of the density function.

Implements sgpp::datadriven::DensityEstimator.

References alpha, and grid.

variance() [2/2]

double sgpp::datadriven::LearnerSGDE::variance ( base::Grid &  grid, base::DataVector &  alpha  )

protected

References sgpp::op_factory::createOperationSecondMoment(), sgpp::base::OperationSecondMoment::doQuadrature(), and mean().

Member Data Documentation

adaptivityConfig

sgpp::base::AdaptivityConfiguration sgpp::datadriven::LearnerSGDE::adaptivityConfig

protected

Referenced by LearnerSGDE(), train(), and trainOnline().

alpha

std::shared_ptr<base::DataVector> sgpp::datadriven::LearnerSGDE::alpha

protected

Referenced by cov(), getSharedSurpluses(), getSurpluses(), initialize(), LearnerSGDE(), mean(), optimizeLambdaCV(), pdf(), python.learner.Classifier.Classifier::refineGrid(), train(), trainOnline(), and variance().

alphas

std::map<int, std::shared_ptr<base::DataVector> > sgpp::datadriven::LearnerSGDE::alphas

protected

Referenced by predict(), storeResults(), and trainOnline().

appearances

std::map<int, size_t> sgpp::datadriven::LearnerSGDE::appearances

protected

Referenced by trainOnline().

avgErrors

base::DataVector sgpp::datadriven::LearnerSGDE::avgErrors

Referenced by trainOnline().

crossvalidationConfig

sgpp::datadriven::CrossvalidationConfiguration sgpp::datadriven::LearnerSGDE::crossvalidationConfig

protected

Referenced by initialize(), LearnerSGDE(), optimizeLambdaCV(), splitset(), and train().

error

double sgpp::datadriven::LearnerSGDE::error

Referenced by LearnerSGDE(), and trainOnline().

grid

std::shared_ptr<base::Grid> sgpp::datadriven::LearnerSGDE::grid

protected

Referenced by cov(), python.uq.learner.Interpolant.Interpolant::doLearningIteration(), python.learner.Classifier.Classifier::evalError(), python.uq.learner.Interpolant.Interpolant::evalError(), python.uq.learner.SimulationLearner.SimulationLearner::getCollocationNodes(), python.uq.learner.SimulationLearner.SimulationLearner::getGrid(), getGrid(), python.uq.learner.SimulationLearner.SimulationLearner::getLearner(), getSharedGrid(), initialize(), python.uq.learner.Regressor.Regressor::learnData(), python.uq.learner.Regressor.Regressor::learnDataWithFolding(), python.uq.learner.Regressor.Regressor::learnDataWithTest(), LearnerSGDE(), mean(), optimizeLambdaCV(), pdf(), python.learner.Classifier.Classifier::refineGrid(), python.learner.Regressor.Regressor::refineGrid(), python.uq.learner.Regressor.Regressor::refineGrid(), python.uq.learner.SimulationLearner.SimulationLearner::refineGrid(), storeResults(), train(), trainOnline(), python.learner.Classifier.Classifier::updateResults(), python.learner.Regressor.Regressor::updateResults(), python.uq.learner.Regressor.Regressor::updateResults(), and variance().

gridConfig

sgpp::base::RegularGridConfiguration sgpp::datadriven::LearnerSGDE::gridConfig

protected

Referenced by createRegularGrid(), getDim(), initialize(), LearnerSGDE(), and trainOnline().

grids

std::map<int, std::shared_ptr<base::Grid> > sgpp::datadriven::LearnerSGDE::grids

protected

Referenced by predict(), storeResults(), and trainOnline().

lambdaReg

double sgpp::datadriven::LearnerSGDE::lambdaReg

protected

Referenced by initialize(), LearnerSGDE(), train(), and trainOnline().

priors

std::map<int, double> sgpp::datadriven::LearnerSGDE::priors

protected

Referenced by predict(), and trainOnline().

regularizationConfig

sgpp::datadriven::RegularizationConfiguration sgpp::datadriven::LearnerSGDE::regularizationConfig

protected

Referenced by computeRegularizationMatrix(), and LearnerSGDE().

solverConfig

sgpp::solver::SLESolverConfiguration sgpp::datadriven::LearnerSGDE::solverConfig

protected

Referenced by LearnerSGDE(), train(), and trainOnline().

trainData

std::shared_ptr<base::DataMatrix> sgpp::datadriven::LearnerSGDE::trainData

protected

Referenced by python.uq.dists.KDEDist.KDEDist::__init__(), getNsamples(), getSamples(), initialize(), LearnerSGDE(), python.uq.dists.KDEDist.KDEDist::marginalize(), python.uq.dists.KDEDist.KDEDist::marginalizeToDimX(), python.uq.dists.KDEDist.KDEDist::toJson(), train(), and trainOnline().

trainLabels

std::shared_ptr<base::DataVector> sgpp::datadriven::LearnerSGDE::trainLabels

protected

Referenced by trainOnline().

usePrior

bool sgpp::datadriven::LearnerSGDE::usePrior

protected

Referenced by LearnerSGDE(), predict(), and trainOnline().

---

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

• datadriven/src/sgpp/datadriven/application/LearnerSGDE.hpp
• datadriven/src/sgpp/datadriven/application/LearnerSGDE.cpp