Top level interface for everything that holds parameters. More...
#include <shark/Core/IParameterizable.h>
Public Member Functions | |
virtual | ~IParameterizable () |
virtual RealVector | parameterVector () const |
Return the parameter vector. More... | |
virtual void | setParameterVector (RealVector const &newParameters) |
Set the parameter vector. More... | |
virtual std::size_t | numberOfParameters () const |
Return the number of parameters. More... | |
Top level interface for everything that holds parameters.
This interface is inherited by AbstractModel for unified access to the parameters of models, but also by objective functions and algorithms with hyper-parameters.
Definition at line 49 of file IParameterizable.h.
|
inlinevirtual |
Definition at line 51 of file IParameterizable.h.
|
inlinevirtual |
Return the number of parameters.
Reimplemented in shark::KernelBudgetedSGDTrainer< InputType, CacheType >, shark::ConcatenatedModel< InputType, OutputType >, shark::AbstractLinearSvmTrainer< InputType >, shark::BinaryLayer, shark::BipolarLayer, shark::AbstractSvmTrainer< InputType, LabelType, Model, Trainer >, shark::AbstractSvmTrainer< InputType, unsigned int, MissingFeaturesKernelExpansion< InputType > >, shark::AbstractSvmTrainer< InputType, unsigned int >, shark::AbstractSvmTrainer< InputType, unsigned int, KernelClassifier< InputType >, AbstractWeightedTrainer< KernelClassifier< InputType > > >, shark::AbstractSvmTrainer< InputType, RealVector, KernelExpansion< InputType > >, shark::GaussianLayer, shark::TruncatedExponentialLayer, shark::KernelSGDTrainer< InputType, CacheType >, shark::KernelExpansion< InputType >, shark::KernelExpansion< RealVector >, shark::ModelKernel< InputType >, shark::WeightedSumKernel< InputType >, shark::Normalizer< DataType >, shark::FFNet< HiddenNeuron, OutputNeuron >, shark::ArgMaxConverter< Model >, shark::ArgMaxConverter< LinearModel< VectorType > >, shark::ArgMaxConverter< KernelExpansion< InputType > >, shark::RNNet, shark::CARTClassifier< LabelType >, shark::CARTClassifier< RealVector >, shark::PolynomialKernel< InputType >, shark::OneClassSvmTrainer< InputType, CacheType >, shark::LinearModel< InputType >, shark::LinearModel< VectorType >, shark::LassoRegression< InputVectorType >, shark::EpsilonSvmTrainer< InputType, CacheType >, shark::SoftNearestNeighborClassifier< InputType >, shark::ThresholdVectorConverter, shark::ConvolutionalRBM< VisibleLayerT, HiddenLayerT, RngT >, shark::ProductKernel< InputType >, shark::CMACMap, shark::NearestNeighborClassifier< InputType >, shark::ConvexCombination, shark::NearestNeighborRegression< InputType >, shark::RBM< VisibleLayerT, HiddenLayerT, RngT >, shark::HierarchicalClustering< InputT >, shark::TiedAutoencoder< HiddenNeuron, OutputNeuron >, shark::Autoencoder< HiddenNeuron, OutputNeuron >, shark::ARDKernelUnconstrained< InputType >, shark::OneVersusOneClassifier< InputType >, shark::NormalizedKernel< InputType >, shark::GaussianRbfKernel< InputType >, shark::OnlineRNNet, shark::Softmax, shark::GaussianNoiseModel, shark::RBFLayer, shark::ImpulseNoiseModel, shark::LDA, shark::ThresholdConverter, shark::DiscreteKernel, shark::MonomialKernel< InputType >, shark::Centroids, shark::LinearRegression, shark::ScaledKernel< InputType >, shark::SigmoidModel, shark::ClusteringModel< InputT, OutputT >, shark::ClusteringModel< InputT, RealVector >, shark::ClusteringModel< InputT, unsigned int >, and shark::LinearNorm.
Definition at line 66 of file IParameterizable.h.
References parameterVector().
Referenced by shark::ProductKernel< InputType >::addKernel(), shark::calculateKernelMatrixParameterDerivative(), shark::RadiusMarginQuotient< InputType, CacheType >::eval(), shark::NegativeGaussianProcessEvidence< InputType, OutputType, LabelType >::eval(), shark::RadiusMarginQuotient< InputType, CacheType >::evalDerivative(), shark::NegativeGaussianProcessEvidence< InputType, OutputType, LabelType >::evalDerivative(), shark::KernelTargetAlignment< InputType, LabelType >::evalDerivative(), shark::initRandomNormal(), shark::initRandomUniform(), shark::ClusteringModel< InputT, unsigned int >::numberOfParameters(), shark::ScaledKernel< InputType >::numberOfParameters(), shark::NormalizedKernel< InputType >::numberOfParameters(), shark::OneClassSvmTrainer< InputType, CacheType >::numberOfParameters(), shark::KernelSGDTrainer< InputType, CacheType >::numberOfParameters(), shark::KernelBudgetedSGDTrainer< InputType, CacheType >::numberOfParameters(), shark::LooErrorCSvm< InputType, CacheType >::numberOfVariables(), shark::NegativeLogLikelihood::numberOfVariables(), shark::RadiusMarginQuotient< InputType, CacheType >::numberOfVariables(), shark::NegativeGaussianProcessEvidence< InputType, OutputType, LabelType >::numberOfVariables(), shark::LooError< ModelTypeT, LabelType >::numberOfVariables(), shark::CrossValidationError< ModelTypeT, LabelTypeT >::numberOfVariables(), shark::KernelTargetAlignment< InputType, LabelType >::numberOfVariables(), shark::OneClassSvmTrainer< InputType, CacheType >::parameterVector(), shark::KernelSGDTrainer< InputType, CacheType >::parameterVector(), shark::KernelBudgetedSGDTrainer< InputType, CacheType >::parameterVector(), shark::RFTrainer::setParameterVector(), shark::OneClassSvmTrainer< InputType, CacheType >::setParameterVector(), shark::KernelSGDTrainer< InputType, CacheType >::setParameterVector(), shark::KernelBudgetedSGDTrainer< InputType, CacheType >::setParameterVector(), shark::NegativeGaussianProcessEvidence< InputType, OutputType, LabelType >::setThreshold(), and shark::CSvmTrainer< InputType, CacheType >::train().
|
inlinevirtual |
Return the parameter vector.
Reimplemented in shark::KernelBudgetedSGDTrainer< InputType, CacheType >, shark::ConcatenatedModel< InputType, OutputType >, shark::AbstractLinearSvmTrainer< InputType >, shark::BinaryLayer, shark::BipolarLayer, shark::GaussianLayer, shark::TruncatedExponentialLayer, shark::AbstractSvmTrainer< InputType, LabelType, Model, Trainer >, shark::AbstractSvmTrainer< InputType, unsigned int, MissingFeaturesKernelExpansion< InputType > >, shark::AbstractSvmTrainer< InputType, unsigned int >, shark::AbstractSvmTrainer< InputType, unsigned int, KernelClassifier< InputType >, AbstractWeightedTrainer< KernelClassifier< InputType > > >, shark::AbstractSvmTrainer< InputType, RealVector, KernelExpansion< InputType > >, shark::KernelSGDTrainer< InputType, CacheType >, shark::ModelKernel< InputType >, shark::FFNet< HiddenNeuron, OutputNeuron >, shark::KernelExpansion< InputType >, shark::KernelExpansion< RealVector >, shark::NBClassifier< InputType, OutputType >, shark::ArgMaxConverter< Model >, shark::ArgMaxConverter< LinearModel< VectorType > >, shark::ArgMaxConverter< KernelExpansion< InputType > >, shark::CARTClassifier< LabelType >, shark::CARTClassifier< RealVector >, shark::Normalizer< DataType >, shark::RNNet, shark::WeightedSumKernel< InputType >, shark::ConvolutionalRBM< VisibleLayerT, HiddenLayerT, RngT >, shark::RBM< VisibleLayerT, HiddenLayerT, RngT >, shark::ThresholdVectorConverter, shark::OneClassSvmTrainer< InputType, CacheType >, shark::TiedAutoencoder< HiddenNeuron, OutputNeuron >, shark::LinearModel< InputType >, shark::LinearModel< VectorType >, shark::Autoencoder< HiddenNeuron, OutputNeuron >, shark::LassoRegression< InputVectorType >, shark::SoftNearestNeighborClassifier< InputType >, shark::EpsilonSvmTrainer< InputType, CacheType >, shark::CMACMap, shark::ProductKernel< InputType >, shark::RFTrainer, shark::HierarchicalClustering< InputT >, shark::NearestNeighborClassifier< InputType >, shark::NearestNeighborRegression< InputType >, shark::ConvexCombination, shark::ARDKernelUnconstrained< InputType >, shark::PolynomialKernel< InputType >, shark::NormalizedKernel< InputType >, shark::OnlineRNNet, shark::RBFLayer, shark::ThresholdConverter, shark::GaussianNoiseModel, shark::Softmax, shark::ImpulseNoiseModel, shark::MonomialKernel< InputType >, shark::LDA, shark::Centroids, shark::OneVersusOneClassifier< InputType >, shark::DiscreteKernel, shark::GaussianRbfKernel< InputType >, shark::LinearRegression, shark::ScaledKernel< InputType >, shark::SigmoidModel, shark::MeanModel< ModelType >, shark::MeanModel< CARTClassifier< RealVector > >, shark::ClusteringModel< InputT, OutputT >, shark::LinearKernel< InputType >, shark::ClusteringModel< InputT, RealVector >, shark::ClusteringModel< InputT, unsigned int >, and shark::LinearNorm.
Definition at line 54 of file IParameterizable.h.
Referenced by numberOfParameters(), shark::ClusteringModel< InputT, unsigned int >::parameterVector(), shark::ScaledKernel< InputType >::parameterVector(), shark::NormalizedKernel< InputType >::parameterVector(), shark::OneClassSvmTrainer< InputType, CacheType >::parameterVector(), shark::NegativeLogLikelihood::proposeStartingPoint(), shark::AbstractMetric< InputType >::write(), and shark::AbstractModel< InputT, unsigned int >::write().
|
inlinevirtual |
Set the parameter vector.
Reimplemented in shark::KernelBudgetedSGDTrainer< InputType, CacheType >, shark::ConcatenatedModel< InputType, OutputType >, shark::AbstractLinearSvmTrainer< InputType >, shark::BinaryLayer, shark::BipolarLayer, shark::AbstractSvmTrainer< InputType, LabelType, Model, Trainer >, shark::AbstractSvmTrainer< InputType, unsigned int, MissingFeaturesKernelExpansion< InputType > >, shark::AbstractSvmTrainer< InputType, unsigned int >, shark::AbstractSvmTrainer< InputType, unsigned int, KernelClassifier< InputType >, AbstractWeightedTrainer< KernelClassifier< InputType > > >, shark::AbstractSvmTrainer< InputType, RealVector, KernelExpansion< InputType > >, shark::GaussianLayer, shark::TruncatedExponentialLayer, shark::KernelSGDTrainer< InputType, CacheType >, shark::ModelKernel< InputType >, shark::FFNet< HiddenNeuron, OutputNeuron >, shark::KernelExpansion< InputType >, shark::KernelExpansion< RealVector >, shark::NBClassifier< InputType, OutputType >, shark::ArgMaxConverter< Model >, shark::ArgMaxConverter< LinearModel< VectorType > >, shark::ArgMaxConverter< KernelExpansion< InputType > >, shark::Normalizer< DataType >, shark::WeightedSumKernel< InputType >, shark::CARTClassifier< LabelType >, shark::CARTClassifier< RealVector >, shark::RNNet, shark::ConvolutionalRBM< VisibleLayerT, HiddenLayerT, RngT >, shark::RBM< VisibleLayerT, HiddenLayerT, RngT >, shark::OneClassSvmTrainer< InputType, CacheType >, shark::LinearModel< InputType >, shark::LinearModel< VectorType >, shark::TiedAutoencoder< HiddenNeuron, OutputNeuron >, shark::Autoencoder< HiddenNeuron, OutputNeuron >, shark::ThresholdVectorConverter, shark::LassoRegression< InputVectorType >, shark::EpsilonSvmTrainer< InputType, CacheType >, shark::SoftNearestNeighborClassifier< InputType >, shark::PolynomialKernel< InputType >, shark::RFTrainer, shark::ProductKernel< InputType >, shark::CMACMap, shark::HierarchicalClustering< InputT >, shark::NearestNeighborClassifier< InputType >, shark::NearestNeighborRegression< InputType >, shark::ConvexCombination, shark::ARDKernelUnconstrained< InputType >, shark::NormalizedKernel< InputType >, shark::OneVersusOneClassifier< InputType >, shark::OnlineRNNet, shark::RBFLayer, shark::GaussianNoiseModel, shark::GaussianRbfKernel< InputType >, shark::ThresholdConverter, shark::Softmax, shark::LDA, shark::ImpulseNoiseModel, shark::DiscreteKernel, shark::MonomialKernel< InputType >, shark::Centroids, shark::LinearRegression, shark::ScaledKernel< InputType >, shark::MeanModel< ModelType >, shark::SigmoidModel, shark::MeanModel< CARTClassifier< RealVector > >, shark::ClusteringModel< InputT, OutputT >, shark::ClusteringModel< InputT, RealVector >, shark::ClusteringModel< InputT, unsigned int >, shark::LinearKernel< InputType >, and shark::LinearNorm.
Definition at line 60 of file IParameterizable.h.
References SHARK_ASSERT.
Referenced by shark::NegativeLogLikelihood::eval(), shark::RadiusMarginQuotient< InputType, CacheType >::eval(), shark::NegativeGaussianProcessEvidence< InputType, OutputType, LabelType >::eval(), shark::CrossValidationError< ModelTypeT, LabelTypeT >::eval(), shark::SvmLogisticInterpretation< InputType >::eval(), shark::LooError< ModelTypeT, LabelType >::eval(), shark::KernelTargetAlignment< InputType, LabelType >::eval(), shark::NegativeLogLikelihood::evalDerivative(), shark::RadiusMarginQuotient< InputType, CacheType >::evalDerivative(), shark::NegativeGaussianProcessEvidence< InputType, OutputType, LabelType >::evalDerivative(), shark::KernelTargetAlignment< InputType, LabelType >::evalDerivative(), shark::SvmLogisticInterpretation< InputType >::evalDerivative(), shark::initRandomNormal(), shark::initRandomUniform(), shark::AbstractMetric< InputType >::read(), shark::AbstractModel< InputT, unsigned int >::read(), shark::ClusteringModel< InputT, unsigned int >::setParameterVector(), shark::ScaledKernel< InputType >::setParameterVector(), shark::NormalizedKernel< InputType >::setParameterVector(), and shark::OneClassSvmTrainer< InputType, CacheType >::setParameterVector().