This class implements a generic decision tree learner. More...

#include <decision_tree_regressor.hpp>

Inheritance diagram for mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >:

[legend]

Collaboration diagram for mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >:

[legend]

Public Types
typedef NumericSplitType< FitnessFunction >	NumericSplit
	Allow access to the numeric split type.

typedef CategoricalSplitType< FitnessFunction >	CategoricalSplit
	Allow access to the categorical split type.

typedef DimensionSelectionType	DimensionSelection
	Allow access to the dimension selection type.

Public Member Functions
	DecisionTreeRegressor ()
	Construct a decision tree without training it. More...

template<typename MatType , typename ResponsesType >
	DecisionTreeRegressor (MatType data, const data::DatasetInfo &datasetInfo, ResponsesType responses, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType())
	Construct the decision tree on the given data and responses, where the data can be both numeric and categorical. More...

template<typename MatType , typename ResponsesType >
	DecisionTreeRegressor (MatType data, ResponsesType responses, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType())
	Construct the decision tree on the given data and responses, assuming that the data is all of the numeric type. More...

template<typename MatType , typename ResponsesType , typename WeightsType >
	DecisionTreeRegressor (MatType data, const data::DatasetInfo &datasetInfo, ResponsesType responses, WeightsType weights, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType(), const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *=0)
	Construct the decision tree on the given data and responses with weights, where the data can be both numeric and categorical. More...

template<typename MatType , typename ResponsesType , typename WeightsType >
	DecisionTreeRegressor (MatType data, ResponsesType responses, WeightsType weights, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType(), const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *=0)
	Construct the decision tree on the given data and responses with weights, assuming that the data is all of the numeric type. More...

template<typename MatType , typename ResponsesType , typename WeightsType >
	DecisionTreeRegressor (const DecisionTreeRegressor &other, MatType data, const data::DatasetInfo &datasetInfo, ResponsesType responses, WeightsType weights, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *=0)
	Take ownership of another decision tree and train on the given data and responses with weights, where the data can be both numeric and categorical. More...

template<typename MatType , typename ResponsesType , typename WeightsType >
	DecisionTreeRegressor (const DecisionTreeRegressor &other, MatType data, ResponsesType responses, WeightsType weights, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType(), const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *=0)
	Take ownership of another decision tree and train on the given data and responses with weights, assuming that the data is all of the numeric type. More...

	DecisionTreeRegressor (const DecisionTreeRegressor &other)
	Copy another tree. More...

	DecisionTreeRegressor (DecisionTreeRegressor &&other)
	Take ownership of another tree. More...

DecisionTreeRegressor &	operator= (const DecisionTreeRegressor &other)
	Copy another tree. More...

DecisionTreeRegressor &	operator= (DecisionTreeRegressor &&other)
	Take ownership of another tree. More...

	~DecisionTreeRegressor ()
	Clean up memory.

template<typename MatType , typename ResponsesType >
double	Train (MatType data, const data::DatasetInfo &datasetInfo, ResponsesType responses, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType())
	Train the decision tree on the given data. More...

template<typename MatType , typename ResponsesType >
double	Train (MatType data, ResponsesType responses, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType())
	Train the decision tree on the given data, assuming that all dimensions are numeric. More...

template<typename MatType , typename ResponsesType , typename WeightsType >
double	Train (MatType data, const data::DatasetInfo &datasetInfo, ResponsesType responses, WeightsType weights, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType(), const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *=0)
	Train the decision tree on the given weighted data. More...

template<typename MatType , typename ResponsesType , typename WeightsType >
double	Train (MatType data, ResponsesType responses, WeightsType weights, const size_t minimumLeafSize=10, const double minimumGainSplit=1e-7, const size_t maximumDepth=0, DimensionSelectionType dimensionSelector=DimensionSelectionType(), const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *=0)
	Train the decision tree on the given weighted data, assuming that all dimensions are numeric. More...

template<typename VecType >
double	Predict (const VecType &point) const
	Make prediction for the given point, using the entire tree. More...

template<typename MatType >
void	Predict (const MatType &data, arma::Row< double > &predictions) const
	Make prediction for the given points, using the entire tree. More...

template<typename Archive >
void	serialize (Archive &ar, const uint32_t)
	Serialize the tree.

size_t	NumChildren () const
	Get the number of children.

size_t	NumLeaves () const
	Get the number of leaves in the tree. More...

const DecisionTreeRegressor &	Child (const size_t i) const
	Get the child of the given index.

DecisionTreeRegressor &	Child (const size_t i)
	Modify the child of the given index (be careful!).

size_t	SplitDimension () const
	Get the split dimension (only meaningful if this is a non-leaf in a trained tree). More...

template<typename VecType >
size_t	CalculateDirection (const VecType &point) const
	Given a point and that this node is not a leaf, calculate the index of the child node this point would go towards. More...

Detailed Description

template<typename FitnessFunction = MSEGain, template< typename > class NumericSplitType = BestBinaryNumericSplit, template< typename > class CategoricalSplitType = AllCategoricalSplit, typename DimensionSelectionType = AllDimensionSelect, bool NoRecursion = false>
class mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >

This class implements a generic decision tree learner.

Its behavior can be controlled via its template arguments.

The class inherits from the auxiliary split information in order to prevent an empty auxiliary split information struct from taking any extra size.

Constructor & Destructor Documentation

◆ DecisionTreeRegressor() [1/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor ( )

Construct a decision tree without training it.

Construct, don't train.

It will be a leaf node.

◆ DecisionTreeRegressor() [2/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType >

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor	(	MatType	data,
		const data::DatasetInfo &	datasetInfo,
		ResponsesType	responses,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`
	)

Construct the decision tree on the given data and responses, where the data can be both numeric and categorical.

Construct and train without weight.

Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data or responses are no longer needed to avoid copies.

Parameters

data	Dataset to train on.
datasetInfo	Type information for each dimension of the dataset.
responses	Responses for each training point.
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

◆ DecisionTreeRegressor() [3/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType >

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor	(	MatType	data,
		ResponsesType	responses,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`
	)

Construct the decision tree on the given data and responses, assuming that the data is all of the numeric type.

Construct and train without weight on numeric data.

Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data or responses are no longer needed to avoid copies.

Parameters

data	Dataset to train on.
responses	Responses for each training point.
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

◆ DecisionTreeRegressor() [4/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType , typename WeightsType >

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor	(	MatType	data,
		const data::DatasetInfo &	datasetInfo,
		ResponsesType	responses,
		WeightsType	weights,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`,
		const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *	= `0`
	)

Construct the decision tree on the given data and responses with weights, where the data can be both numeric and categorical.

Construct and train with weights.

Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data, responses or weights are no longer needed to avoid copies.

Parameters

data	Dataset to train on.
datasetInfo	Type information for each dimension of the dataset.
responses	Responses for each training point.
weights	The weight list of given label.
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

◆ DecisionTreeRegressor() [5/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType , typename WeightsType >

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor	(	MatType	data,
		ResponsesType	responses,
		WeightsType	weights,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`,
		const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *	= `0`
	)

Construct the decision tree on the given data and responses with weights, assuming that the data is all of the numeric type.

Construct and train on numeric data with weights.

Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data, responses or weights are no longer needed to avoid copies.

Parameters

data	Dataset to train on.
responses	Responses for each training point.
weights	The Weight list of given labels.
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

◆ DecisionTreeRegressor() [6/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType , typename WeightsType >

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor	(	const DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion > &	other,
		MatType	data,
		const data::DatasetInfo &	datasetInfo,
		ResponsesType	responses,
		WeightsType	weights,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *	= `0`
	)

Take ownership of another decision tree and train on the given data and responses with weights, where the data can be both numeric and categorical.

Take ownership of another tree and train with weights.

Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data, responses or weights are no longer needed to avoid copies.

Parameters

other	Tree to take ownership of.
data	Dataset to train on.
datasetInfo	Type information for each dimension of the dataset.
responses	Responses for each training point.
weights	The weight list of given label.
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.

◆ DecisionTreeRegressor() [7/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType , typename WeightsType >

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor	(	const DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion > &	other,
		MatType	data,
		ResponsesType	responses,
		WeightsType	weights,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`,
		const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *	= `0`
	)

Take ownership of another decision tree and train on the given data and responses with weights, assuming that the data is all of the numeric type.

Take ownership of another tree and train with weights.

Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data, responses or weights are no longer needed to avoid copies.

Parameters

other	Tree to take ownership of.
data	Dataset to train on.
responses	Responses for each training point.
weights	The Weight list of given labels.
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

◆ DecisionTreeRegressor() [8/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor ( const DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion > & other )

Copy another tree.

This may use a lot of memory—be sure that it's what you want to do.

Parameters

other Tree to copy.

◆ DecisionTreeRegressor() [9/9]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::DecisionTreeRegressor ( DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion > && other )

Take ownership of another tree.

Parameters

other Tree to take ownership of.

Member Function Documentation

◆ CalculateDirection()

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename VecType >

size_t mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::CalculateDirection ( const VecType & point ) const

Given a point and that this node is not a leaf, calculate the index of the child node this point would go towards.

This method is primarily used by the Predict() function, but it can be used in a standalone sense too.

Parameters

point Point to predict.

◆ NumLeaves()

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

size_t mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::NumLeaves ( ) const

Get the number of leaves in the tree.

Return the number of leaves.

◆ operator=() [1/2]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion > & mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::operator= ( const DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion > & other )

Copy another tree.

This may use a lot of memory—be sure that it's what you want to do.

Parameters

other Tree to copy.

◆ operator=() [2/2]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion > & mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::operator= ( DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion > && other )

Take ownership of another tree.

Parameters

other Tree to take ownership of.

◆ Predict() [1/2]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename VecType >

double mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::Predict ( const VecType & point ) const

Make prediction for the given point, using the entire tree.

Return the prediction.

The predicted label is returned.

Parameters

point Point to predict.

◆ Predict() [2/2]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType >

void mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::Predict	(	const MatType &	data,
		arma::Row< double > &	predictions
	)		const

Make prediction for the given points, using the entire tree.

Return the predictions for a set of points.

The predicted responses for each point are stored in the given vector.

Parameters

data	Set of points to predict.
predictions	This will be filled with predictions for each point.

◆ SplitDimension()

template<typename FitnessFunction = MSEGain, template< typename > class NumericSplitType = BestBinaryNumericSplit, template< typename > class CategoricalSplitType = AllCategoricalSplit, typename DimensionSelectionType = AllDimensionSelect, bool NoRecursion = false>

size_t mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::SplitDimension ( ) const

inline

Get the split dimension (only meaningful if this is a non-leaf in a trained tree).

◆ Train() [1/4]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType >

double mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::Train	(	MatType	data,
		const data::DatasetInfo &	datasetInfo,
		ResponsesType	responses,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`
	)

Train the decision tree on the given data.

Train on the given data.

This will overwrite the existing model. The data may have numeric and categorical types, specified by the datasetInfo parameter. Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data or responses are no longer needed to avoid copies.

Parameters

data	Dataset to train on.
datasetInfo	Type information for each dimension.
responses	Responses for each training point.
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

Returns: The final entropy of decision tree.

◆ Train() [2/4]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType >

double mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::Train	(	MatType	data,
		ResponsesType	responses,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`
	)

Train the decision tree on the given data, assuming that all dimensions are numeric.

Train on the given data, assuming all dimensions are numeric.

This will overwrite the given model. Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data or responses are no longer needed to avoid copies.

Parameters

data	Dataset to train on.
responses	Responses for each training point.
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

Returns: The final entropy of decision tree.

◆ Train() [3/4]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType , typename WeightsType >

double mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::Train	(	MatType	data,
		const data::DatasetInfo &	datasetInfo,
		ResponsesType	responses,
		WeightsType	weights,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`,
		const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *	= `0`
	)

Train the decision tree on the given weighted data.

Train on the given weighted data.

This will overwrite the existing model. The data may have numeric and categorical types, specified by the datasetInfo parameter. Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data, responses or weights are no longer needed to avoid copies.

Parameters

data	Dataset to train on.
datasetInfo	Type information for each dimension.
responses	Responses for each training point.
weights	Weights of all the labels
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

Returns: The final entropy of decision tree.

◆ Train() [4/4]

template<typename FitnessFunction , template< typename > class NumericSplitType, template< typename > class CategoricalSplitType, typename DimensionSelectionType , bool NoRecursion>

template<typename MatType , typename ResponsesType , typename WeightsType >

double mlpack::tree::DecisionTreeRegressor< FitnessFunction, NumericSplitType, CategoricalSplitType, DimensionSelectionType, NoRecursion >::Train	(	MatType	data,
		ResponsesType	responses,
		WeightsType	weights,
		const size_t	minimumLeafSize = `10`,
		const double	minimumGainSplit = `1e-7`,
		const size_t	maximumDepth = `0`,
		DimensionSelectionType	dimensionSelector = `DimensionSelectionType()`,
		const std::enable_if_t< arma::is_arma_type< typename std::remove_reference< WeightsType >::type >::value > *	= `0`
	)

Train the decision tree on the given weighted data, assuming that all dimensions are numeric.

Train on the given weighted all numeric data.

This will overwrite the given model. Setting minimumLeafSize and minimumGainSplit too small may cause the tree to overfit, but setting them too large may cause it to underfit.

Use std::move if data, responses or weights are no longer needed to avoid copies.

Parameters

data	Dataset to train on.
responses	Responses for each training point.
weights	Weights of all the labels
minimumLeafSize	Minimum number of points in each leaf node.
minimumGainSplit	Minimum gain for the node to split.
maximumDepth	Maximum depth for the tree.
dimensionSelector	Instantiated dimension selection policy.

Returns: The final entropy of decision tree.

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

src/mlpack/methods/decision_tree/decision_tree_regressor.hpp
src/mlpack/methods/decision_tree/decision_tree_regressor_impl.hpp

Public Types

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ DecisionTreeRegressor() [1/9]

◆ DecisionTreeRegressor() [2/9]

◆ DecisionTreeRegressor() [3/9]

◆ DecisionTreeRegressor() [4/9]

◆ DecisionTreeRegressor() [5/9]

◆ DecisionTreeRegressor() [6/9]

◆ DecisionTreeRegressor() [7/9]

◆ DecisionTreeRegressor() [8/9]

◆ DecisionTreeRegressor() [9/9]

Member Function Documentation

◆ CalculateDirection()

◆ NumLeaves()

◆ operator=() [1/2]

◆ operator=() [2/2]

◆ Predict() [1/2]

◆ Predict() [2/2]

◆ SplitDimension()

◆ Train() [1/4]

◆ Train() [2/4]

◆ Train() [3/4]

◆ Train() [4/4]