freeling::mlDTree Class Reference

Class mlDTree implements a multilabel decision tree that can be used by adaboost as a weak rules. More...

#include <weakrule.h>

Inheritance diagram for freeling::mlDTree:

Collaboration diagram for freeling::mlDTree:

List of all members.

Public Member Functions
	mlDTree (mlDTree_params *p)
void	classify (const example &i, double pred[])
	Classification Important: pred is an array of predictions, one for each label the function *adds* its predicion for each label.
void	write_to_stream (std::wostream *os)
	I/O operations.
void	read_from_stream (std::wistream *is)
	I/O operations.
void	learn (const dataset &ds, double &Z)
	Learning.
Private Member Functions
void	classify (const example &i, double pred[], tree< dt_node >::iterator t)
	auxiliar classifying function
void	write_to_stream (tree< dt_node >::iterator t, std::wostream *os)
	auxiliar I/O functions
tree< dt_node >	read_dt (std::wistream *is)
tree< dt_node >	learn (const dataset &ds, double &Z, int depth)
	auxiliar learning functions
bool	stopping_criterion (const dataset &ds, int depth)
int	best_feature (const dataset &ds, double *W)
	W is W[2][nlabels][2].
void	Cprediction (int v, double *W, double result[])
	W is W[v][nlabels][2]; result is result[nlabels].
double	Zcalculus (double *W, int ndim)
	We will re-use Z computed during decision-tree building so we ignore default Zcalculus and do it our way.
	mlDTree (const mlDTree &wr0)
	copy constructor forbidden
Private Attributes
mlDTree_params	params
tree< dt_node >	rule
std::set< int >	used_features

---

Detailed Description

Class mlDTree implements a multilabel decision tree that can be used by adaboost as a weak rules.

---

Constructor & Destructor Documentation

freeling::mlDTree::mlDTree

(

const mlDTree &

wr0

)

[private]

copy constructor forbidden

freeling::mlDTree::mlDTree

(

mlDTree_params *

p

)

---

Member Function Documentation

int freeling::mlDTree::best_feature	(	const dataset &	ds,
		double *	W
	)		[private]

W is W[2][nlabels][2].

void freeling::mlDTree::classify	(	const example &	i,
		double	pred[],
		tree< dt_node >::iterator	t
	)		[private]

auxiliar classifying function

void freeling::mlDTree::classify	(	const example &	i,
		double	pred[]
	)		[virtual]

Classification Important: pred is an array of predictions, one for each label the function *adds* its predicion for each label.

Implements freeling::weak_rule.

void freeling::mlDTree::Cprediction	(	int	v,
		double *	W,
		double	result[]
	)		[private]

W is W[v][nlabels][2]; result is result[nlabels].

tree<dt_node> freeling::mlDTree::learn	(	const dataset &	ds,
		double &	Z,
		int	depth
	)		[private]

auxiliar learning functions

void freeling::mlDTree::learn	(	const dataset &	ds,
		double &	Z
	)		[virtual]

Learning.

Implements freeling::weak_rule.

tree<dt_node> freeling::mlDTree::read_dt

(

std::wistream *

is

)

[private]

void freeling::mlDTree::read_from_stream

(

std::wistream *

is

)

[virtual]

I/O operations.

Implements freeling::weak_rule.

bool freeling::mlDTree::stopping_criterion	(	const dataset &	ds,
		int	depth
	)		[private]

void freeling::mlDTree::write_to_stream	(	tree< dt_node >::iterator	t,
		std::wostream *	os
	)		[private]

auxiliar I/O functions

void freeling::mlDTree::write_to_stream

(

std::wostream *

os

)

[virtual]

I/O operations.

Implements freeling::weak_rule.

double freeling::mlDTree::Zcalculus	(	double *	W,
		int	ndim
	)		[private]

We will re-use Z computed during decision-tree building so we ignore default Zcalculus and do it our way.

W is W[ndim][nlabels][2]

---

Member Data Documentation

mlDTree_params freeling::mlDTree::params [private]

tree<dt_node> freeling::mlDTree::rule [private]

std::set<int> freeling::mlDTree::used_features [private]

---

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

• weakrule.h