Top level interface for everything that holds parameters. More...
#include <shark/Core/IParameterizable.h>
Inheritance diagram for shark::IParameterizable< VectorType >:
Public Types | |
| typedef VectorType | ParameterVectorType |
Public Member Functions | |
| virtual | ~IParameterizable () |
| virtual ParameterVectorType | parameterVector () const |
| Return the parameter vector. More... | |
| virtual void | setParameterVector (ParameterVectorType const &newParameters) |
| Set the parameter vector. More... | |
| virtual std::size_t | numberOfParameters () const |
| Return the number of parameters. More... | |
Detailed Description
template<class VectorType = RealVector>
class shark::IParameterizable< VectorType >
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.
the type of parameter vector can be chosen, e.g. to change precision or port parameters to GPU
Definition at line 53 of file IParameterizable.h.
Member Typedef Documentation
◆ ParameterVectorType
| typedef VectorType shark::IParameterizable< VectorType >::ParameterVectorType |
Definition at line 55 of file IParameterizable.h.
Constructor & Destructor Documentation
◆ ~IParameterizable()
|
inlinevirtual |
Definition at line 56 of file IParameterizable.h.
Member Function Documentation
◆ numberOfParameters()
|
inlinevirtual |
Return the number of parameters.
Reimplemented in shark::KernelBudgetedSGDTrainer< InputType, CacheType >, 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::AbstractSvmTrainer< InputType, unsigned int, KernelExpansion< InputType > >, shark::GaussianLayer, shark::TruncatedExponentialLayer, shark::NeuronLayer< NeuronType, VectorType >, shark::KernelSGDTrainer< InputType, CacheType >, shark::KernelExpansion< InputType >, shark::ModelKernel< InputType >, shark::WeightedSumKernel< InputType >, shark::GaussianTaskKernel< InputTypeT >, shark::LinearSAGTrainer< InputType, LabelType >, shark::PolynomialKernel< InputType >, shark::OneClassSvmTrainer< InputType, CacheType >, shark::LinearModel< InputType, ActivationFunction >, shark::LinearModel< VectorType >, shark::Conv2DModel< VectorType, ActivationFunction >, shark::LogisticRegression< InputVectorType >, shark::ConcatenatedModel< VectorType >, shark::ProductKernel< InputType >, shark::ProductKernel< MultiTaskSample< InputTypeT > >, shark::LassoRegression< InputVectorType >, shark::EpsilonSvmTrainer< InputType, CacheType >, shark::Normalizer< VectorType >, shark::HierarchicalClustering< InputT >, shark::CMACMap, shark::RBM< VisibleLayerT, HiddenLayerT, randomT >, shark::CARTree< LabelType >, shark::CARTree< unsigned int >, shark::ARDKernelUnconstrained< InputType >, shark::OneVersusOneClassifier< InputType, VectorType >, shark::NormalizedKernel< InputType >, shark::GaussianRbfKernel< InputType >, shark::RBFLayer, shark::LDA, shark::DiscreteKernel, shark::Classifier< Model >, shark::MonomialKernel< InputType >, shark::Classifier< detail::BaseNearestNeighbor< InputType, unsigned int > >, shark::Classifier< LinearModel< VectorType > >, shark::Classifier< MeanModel< CARTree< unsigned int > > >, shark::Classifier< KernelExpansion< InputType > >, shark::Centroids, shark::PointSetKernel< InputType >, shark::LinearRegression, shark::ScaledKernel< InputType >, shark::DropoutLayer< VectorType >, shark::ClusteringModel< InputT, OutputT >, shark::ClusteringModel< InputT, RealVector >, and shark::ClusteringModel< InputT, unsigned int >.
Definition at line 71 of file IParameterizable.h.
Referenced by shark::ProductKernel< MultiTaskSample< InputTypeT > >::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::ClusteringModel< InputT, unsigned int >::numberOfParameters(), shark::ScaledKernel< InputType >::numberOfParameters(), shark::PointSetKernel< 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::VariationalAutoencoderError::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::OneClassSvmTrainer< InputType, CacheType >::setParameterVector(), shark::KernelSGDTrainer< InputType, CacheType >::setParameterVector(), shark::KernelBudgetedSGDTrainer< InputType, CacheType >::setParameterVector(), and shark::NegativeGaussianProcessEvidence< InputType, OutputType, LabelType >::setThreshold().
◆ parameterVector()
|
inlinevirtual |
Return the parameter vector.
Reimplemented in shark::KernelBudgetedSGDTrainer< InputType, CacheType >, 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::AbstractSvmTrainer< InputType, unsigned int, KernelExpansion< InputType > >, shark::GaussianLayer, shark::TruncatedExponentialLayer, shark::NeuronLayer< NeuronType, VectorType >, shark::RFTrainer< RealVector >, shark::KernelSGDTrainer< InputType, CacheType >, shark::ModelKernel< InputType >, shark::KernelExpansion< InputType >, shark::GaussianTaskKernel< InputTypeT >, shark::WeightedSumKernel< InputType >, shark::LinearSAGTrainer< InputType, LabelType >, shark::RBM< VisibleLayerT, HiddenLayerT, randomT >, shark::LinearModel< InputType, ActivationFunction >, shark::LinearModel< VectorType >, shark::LogisticRegression< InputVectorType >, shark::OneClassSvmTrainer< InputType, CacheType >, shark::Conv2DModel< VectorType, ActivationFunction >, shark::RFTrainer< unsigned int >, shark::LassoRegression< InputVectorType >, shark::CMACMap, shark::CARTree< LabelType >, shark::CARTree< unsigned int >, shark::EpsilonSvmTrainer< InputType, CacheType >, shark::ConcatenatedModel< VectorType >, shark::ProductKernel< InputType >, shark::ProductKernel< MultiTaskSample< InputTypeT > >, shark::HierarchicalClustering< InputT >, shark::Normalizer< VectorType >, shark::PolynomialKernel< InputType >, shark::MeanModel< ModelType >, shark::MeanModel< CARTree< LabelType > >, shark::MeanModel< CARTree< unsigned int > >, shark::ARDKernelUnconstrained< InputType >, shark::NormalizedKernel< InputType >, shark::RBFLayer, shark::MonomialKernel< InputType >, shark::LDA, shark::Centroids, shark::OneVersusOneClassifier< InputType, VectorType >, shark::Classifier< Model >, shark::DiscreteKernel, shark::GaussianRbfKernel< InputType >, shark::Classifier< detail::BaseNearestNeighbor< InputType, unsigned int > >, shark::Classifier< LinearModel< VectorType > >, shark::Classifier< MeanModel< CARTree< unsigned int > > >, shark::Classifier< KernelExpansion< InputType > >, shark::PointSetKernel< InputType >, shark::LinearRegression, shark::ScaledKernel< InputType >, shark::DropoutLayer< VectorType >, shark::ClusteringModel< InputT, OutputT >, shark::LinearKernel< InputType >, shark::ClusteringModel< InputT, RealVector >, and shark::ClusteringModel< InputT, unsigned int >.
Definition at line 59 of file IParameterizable.h.
Referenced by shark::IParameterizable< ParameterType >::numberOfParameters(), shark::ClusteringModel< InputT, unsigned int >::parameterVector(), shark::ScaledKernel< InputType >::parameterVector(), shark::PointSetKernel< InputType >::parameterVector(), shark::NormalizedKernel< InputType >::parameterVector(), shark::KernelSGDTrainer< InputType, CacheType >::parameterVector(), shark::KernelBudgetedSGDTrainer< InputType, CacheType >::parameterVector(), shark::NegativeLogLikelihood::proposeStartingPoint(), and shark::VariationalAutoencoderError::proposeStartingPoint().
◆ setParameterVector()
|
inlinevirtual |
Set the parameter vector.
Reimplemented in shark::KernelBudgetedSGDTrainer< InputType, CacheType >, 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::AbstractSvmTrainer< InputType, unsigned int, KernelExpansion< InputType > >, shark::GaussianLayer, shark::TruncatedExponentialLayer, shark::NeuronLayer< NeuronType, VectorType >, shark::KernelSGDTrainer< InputType, CacheType >, shark::RFTrainer< RealVector >, shark::ModelKernel< InputType >, shark::KernelExpansion< InputType >, shark::WeightedSumKernel< InputType >, shark::GaussianTaskKernel< InputTypeT >, shark::LinearSAGTrainer< InputType, LabelType >, shark::RBM< VisibleLayerT, HiddenLayerT, randomT >, shark::OneClassSvmTrainer< InputType, CacheType >, shark::LinearModel< VectorType >, shark::LogisticRegression< InputVectorType >, shark::Conv2DModel< VectorType, ActivationFunction >, shark::ConcatenatedModel< VectorType >, shark::LassoRegression< InputVectorType >, shark::ProductKernel< InputType >, shark::ProductKernel< MultiTaskSample< InputTypeT > >, shark::RFTrainer< unsigned int >, shark::PolynomialKernel< InputType >, shark::EpsilonSvmTrainer< InputType, CacheType >, shark::CARTree< LabelType >, shark::CARTree< unsigned int >, shark::CMACMap, shark::HierarchicalClustering< InputT >, shark::Normalizer< VectorType >, shark::MeanModel< CARTree< LabelType > >, shark::MeanModel< CARTree< unsigned int > >, shark::ARDKernelUnconstrained< InputType >, shark::NormalizedKernel< InputType >, shark::OneVersusOneClassifier< InputType, VectorType >, shark::RBFLayer, shark::GaussianRbfKernel< InputType >, shark::LDA, shark::DiscreteKernel, shark::MonomialKernel< InputType >, shark::Classifier< Model >, shark::Centroids, shark::Classifier< detail::BaseNearestNeighbor< InputType, unsigned int > >, shark::Classifier< LinearModel< VectorType > >, shark::Classifier< MeanModel< CARTree< unsigned int > > >, shark::Classifier< KernelExpansion< InputType > >, shark::PointSetKernel< InputType >, shark::LinearRegression, shark::ScaledKernel< InputType >, shark::DropoutLayer< VectorType >, shark::ClusteringModel< InputT, OutputT >, shark::ClusteringModel< InputT, RealVector >, shark::ClusteringModel< InputT, unsigned int >, and shark::LinearKernel< InputType >.
Definition at line 65 of file IParameterizable.h.
Referenced by shark::NegativeLogLikelihood::eval(), shark::LooErrorCSvm< InputType, CacheType >::eval(), shark::VariationalAutoencoderError::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::VariationalAutoencoderError::evalDerivative(), shark::NegativeGaussianProcessEvidence< InputType, OutputType, LabelType >::evalDerivative(), shark::SvmLogisticInterpretation< InputType >::evalDerivative(), shark::KernelTargetAlignment< InputType, LabelType >::evalDerivative(), shark::ClusteringModel< InputT, unsigned int >::setParameterVector(), shark::ScaledKernel< InputType >::setParameterVector(), shark::PointSetKernel< InputType >::setParameterVector(), shark::NormalizedKernel< InputType >::setParameterVector(), shark::OneClassSvmTrainer< InputType, CacheType >::setParameterVector(), shark::KernelSGDTrainer< InputType, CacheType >::setParameterVector(), and shark::KernelBudgetedSGDTrainer< InputType, CacheType >::setParameterVector().
The documentation for this class was generated from the following file:
- include/shark/Core/IParameterizable.h