mlpack::tree::HoeffdingCategoricalSplit< FitnessFunction > Class Template Reference

This is the standard Hoeffding-bound categorical feature proposed in the paper below: More...

#include <hoeffding_categorical_split.hpp>

Public Types
typedef CategoricalSplitInfo	SplitInfo
	The type of split information required by the HoeffdingCategoricalSplit.

Public Member Functions
	HoeffdingCategoricalSplit (const size_t numCategories=0, const size_t numClasses=0)
	Create the HoeffdingCategoricalSplit given a number of categories for this dimension and a number of classes. More...
	HoeffdingCategoricalSplit (const size_t numCategories, const size_t numClasses, const HoeffdingCategoricalSplit &other)
	Create the HoeffdingCategoricalSplit given a number of categories for this dimension and a number of classes and another HoeffdingCategoricalSplit to take parameters from. More...
template<typename eT >
void	Train (eT value, const size_t label)
	Train on the given value with the given label. More...
void	EvaluateFitnessFunction (double &bestFitness, double &secondBestFitness) const
	Given the points seen so far, evaluate the fitness function, returning the gain for the best possible split and the second best possible split. More...
size_t	NumChildren () const
	Return the number of children, if the node were to split.
void	Split (arma::Col< size_t > &childMajorities, SplitInfo &splitInfo)
	Gather the information for a split: get the labels of the child majorities, and initialize the SplitInfo object. More...
size_t	MajorityClass () const
	Get the majority class seen so far.
double	MajorityProbability () const
	Get the probability of the majority class given the points seen so far.
template<typename Archive >
void	serialize (Archive &ar, const uint32_t)
	Serialize the categorical split.

Detailed Description

template<typename FitnessFunction>
class mlpack::tree::HoeffdingCategoricalSplit< FitnessFunction >

This is the standard Hoeffding-bound categorical feature proposed in the paper below:

@inproceedings{domingos2000mining,
  title={{Mining High-Speed Data Streams}},
  author={Domingos, P. and Hulten, G.},
  year={2000},
  booktitle={Proceedings of the Sixth ACM SIGKDD International Conference on
      Knowledge Discovery and Data Mining (KDD '00)},
  pages={71--80}
}

This class will track the sufficient statistics of the training points it has seen. The HoeffdingSplit class (and other related classes) can use this class to track categorical features and split decision tree nodes.

Template Parameters

FitnessFunction

Fitness function to use for calculating gain.

Constructor & Destructor Documentation

HoeffdingCategoricalSplit() [1/2]

template<typename FitnessFunction >

mlpack::tree::HoeffdingCategoricalSplit< FitnessFunction >::HoeffdingCategoricalSplit	(	const size_t	numCategories = 0,
		const size_t	numClasses = 0
	)

Create the HoeffdingCategoricalSplit given a number of categories for this dimension and a number of classes.

Parameters

numCategories	Number of categories in this dimension.
numClasses	Number of classes in this dimension.

HoeffdingCategoricalSplit() [2/2]

template<typename FitnessFunction >

mlpack::tree::HoeffdingCategoricalSplit< FitnessFunction >::HoeffdingCategoricalSplit	(	const size_t	numCategories,
		const size_t	numClasses,
		const HoeffdingCategoricalSplit< FitnessFunction > &	other
	)

Create the HoeffdingCategoricalSplit given a number of categories for this dimension and a number of classes and another HoeffdingCategoricalSplit to take parameters from.

In this particular case, there are no parameters to take, but this constructor is required by the HoeffdingTree class.

Member Function Documentation

EvaluateFitnessFunction()

template<typename FitnessFunction >

void mlpack::tree::HoeffdingCategoricalSplit< FitnessFunction >::EvaluateFitnessFunction	(	double &	bestFitness,
		double &	secondBestFitness
	)		const

Given the points seen so far, evaluate the fitness function, returning the gain for the best possible split and the second best possible split.

In this splitting technique, we only split one possible way, so secondBestFitness will always be 0.

Parameters

bestFitness	The fitness function result for this split.
secondBestFitness	This is always set to 0 (this split only splits one way).

Split()

template<typename FitnessFunction >

void mlpack::tree::HoeffdingCategoricalSplit< FitnessFunction >::Split	(	arma::Col< size_t > &	childMajorities,
		SplitInfo &	splitInfo
	)

Gather the information for a split: get the labels of the child majorities, and initialize the SplitInfo object.

Parameters

childMajorities	Majorities of child nodes to be created.
splitInfo	Information for splitting.

Train()

template<typename FitnessFunction >

template<typename eT >

void mlpack::tree::HoeffdingCategoricalSplit< FitnessFunction >::Train	(	eT	value,
		const size_t	label
	)

Train on the given value with the given label.

Parameters

value	Value to train on.
label	Label to train on.

---

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

• src/mlpack/methods/hoeffding_trees/hoeffding_categorical_split.hpp
• src/mlpack/methods/hoeffding_trees/hoeffding_categorical_split_impl.hpp