Ordinal_BLE Class Reference

Ordinal regression via binary learning on extended examples. More...

#include <ordinal_ble.h>

Inheritance diagram for Ordinal_BLE:

[legend]Collaboration diagram for Ordinal_BLE:


Public Member Functions
	Ordinal_BLE ()
	Ordinal_BLE (const Ordinal_BLE &)
const Ordinal_BLE &	operator= (const Ordinal_BLE &)
	Ordinal_BLE (std::istream &is)
virtual const id_t &	id () const
virtual Ordinal_BLE *	create () const
	Create a new object using the default constructor.
virtual Ordinal_BLE *	clone () const
	Create a new object by replicating itself.
void	set_model (const LearnModel &)
	set the underlying learning model
const LearnModel &	model () const
	the underlying model
UINT	n_rank () const
	the number of ranks from the training set
virtual bool	support_weighted_data () const
	Whether the learning model/algorithm supports unequally weighted data.
virtual REAL	c_error (const Output &out, const Output &y) const
	Error measure for classification problems.
virtual REAL	r_error (const Output &out, const Output &y) const
	Error measure for regression problems.
virtual void	set_train_data (const pDataSet &, const pDataWgt &=0)
	Set the data set and sample weight to be used in training.
virtual void	train ()
	Train with preset data set and sample weight.
virtual void	reset ()
virtual Output	operator() (const Input &) const

bool	full_extension () const
void	set_full_extension (bool=true)
const ECOC_TABLE &	ECOC_table () const
const EXT_TABLE &	extension_table () const
void	set_tables (const ECOC_TABLE &, const EXT_TABLE &)
void	set_tables (BLE_TYPE, UINT)
Protected Member Functions
void	extend_input (const Input &, UINT, Input &) const
void	extend_example (const Input &, UINT, UINT, Input &, REAL &) const
void	extend_data ()
virtual REAL	ECOC_distance (const Output &, const ECOC_VECTOR &) const
const std::vector< REAL > &	distances (const Input &) const
virtual bool	serialize (std::ostream &, ver_list &) const
virtual bool	unserialize (std::istream &, ver_list &, const id_t &=NIL_ID)
Protected Attributes
pLearnModel	lm
	the learning model
bool	full_ext
	use the full extension or the partial one?
ECOC_TABLE	out_tab
	K (nrank) by T (n_hyp) output ECC matrix.
EXT_TABLE	ext_tab
	T (n_hyp) by E (n_ext) extension matrix.
UINT	n_ext
pDataSet	ext_d
	the extended dataset
pDataWgt	ext_w
	the weight for the extended set
UINT	d_nrank
	number of ranking levels of the training set
bool	reset_data
	whether to reset the training set for lm
std::vector< REAL >	local_d

Detailed Description

Ordinal regression via binary learning on extended examples.

Definition at line 23 of file ordinal_ble.h.

Constructor & Destructor Documentation

Ordinal_BLE ( ) [inline]

Definition at line 32 of file ordinal_ble.h.
References LearnModel::set_dimensions().
Referenced by Ordinal_BLE::clone(), and Ordinal_BLE::create().

Ordinal_BLE ( const Ordinal_BLE & )

Definition at line 78 of file ordinal_ble.cpp.
References Ordinal_BLE::lm.

Ordinal_BLE ( std::istream & is ) [inline, explicit]

Definition at line 36 of file ordinal_ble.h.

Member Function Documentation

REAL c_error ( const Output & out,

const Output & y

) const [virtual]

Error measure for classification problems.

Parameters:

out is the output from the learned hypothesis.

y is the real output.

Returns:
Classification error between out and y. The error measure is not necessary symmetric. A commonly used measure is out != y.

Reimplemented from LearnModel.
Definition at line 152 of file ordinal_ble.cpp.
References LearnModel::n_output(), OUT2RANK, and VALIDRANK.

virtual Ordinal_BLE* clone ( ) const [inline, virtual]

Create a new object by replicating itself.

Returns:
A pointer to the new copy.
The code for a derived class Derived is always
return new Derived(*this);
Though seemingly redundant, it helps to copy an object without knowing the real type of the object.
See also:
C++ FAQ Lite 20.6

Implements LearnModel.
Definition at line 40 of file ordinal_ble.h.
References Ordinal_BLE::Ordinal_BLE().

virtual Ordinal_BLE* create ( ) const [inline, virtual]

Create a new object using the default constructor.
The code for a derived class Derived is always
return new Derived();

Implements LearnModel.
Definition at line 39 of file ordinal_ble.h.
References Ordinal_BLE::Ordinal_BLE().

const std::vector< REAL > & distances ( const Input & ) const [protected]

Definition at line 304 of file ordinal_ble.cpp.
References Ordinal_BLE::extend_input(), and Ordinal_BLE::lm.
Referenced by Ordinal_BLE::operator()().

REAL ECOC_distance ( const Output & ,

const ECOC_VECTOR &

) const [protected, virtual]

Definition at line 295 of file ordinal_ble.cpp.

const ECOC_TABLE& ECOC_table ( ) const [inline]

Definition at line 50 of file ordinal_ble.h.
References Ordinal_BLE::out_tab.

void extend_data ( ) [protected]

Definition at line 203 of file ordinal_ble.cpp.
References dataset::append(), Ordinal_BLE::extend_example(), LearnModel::n_samples, nrank, OUT2RANK, and LearnModel::ptd.
Referenced by Ordinal_BLE::train().

void extend_example ( const Input & ,

UINT ,

UINT ,

Input & ,

REAL &

) const [protected]

Definition at line 287 of file ordinal_ble.cpp.
References Ordinal_BLE::extend_input(), nrank, and Ordinal_BLE::out_tab.
Referenced by Ordinal_BLE::extend_data().

void extend_input ( const Input & x,

UINT t,

Input & ext_x

) const [protected]

t: hypothesis index

Definition at line 278 of file ordinal_ble.cpp.
References Ordinal_BLE::ext_tab.
Referenced by Ordinal_BLE::distances(), and Ordinal_BLE::extend_example().

const EXT_TABLE& extension_table ( ) const [inline]

Definition at line 51 of file ordinal_ble.h.
References Ordinal_BLE::ext_tab.

bool full_extension ( ) const [inline]

Definition at line 48 of file ordinal_ble.h.
References Ordinal_BLE::full_ext.

virtual const id_t& id ( ) const [virtual]

Returns:
Class ID string (class name)

Implements Object.

const LearnModel& model ( ) const [inline]

the underlying model

Definition at line 45 of file ordinal_ble.h.
References Ordinal_BLE::lm.

UINT n_rank ( ) const [inline]

the number of ranks from the training set

Definition at line 57 of file ordinal_ble.h.
References Ordinal_BLE::d_nrank.

Output operator() ( const Input & ) const [virtual]

Implements LearnModel.
Definition at line 270 of file ordinal_ble.cpp.
References Ordinal_BLE::distances(), GET_BEST_RANK, RANK2OUT, and LearnModel::valid_dimensions().

const Ordinal_BLE & operator= ( const Ordinal_BLE & )

Definition at line 87 of file ordinal_ble.cpp.
References Ordinal_BLE::d_nrank, Ordinal_BLE::ext_d, Ordinal_BLE::ext_tab, Ordinal_BLE::ext_w, Ordinal_BLE::full_ext, Ordinal_BLE::lm, Ordinal_BLE::n_ext, Ordinal_BLE::out_tab, and Ordinal_BLE::reset_data.

REAL r_error ( const Output & out,

const Output & y

) const [virtual]

Error measure for regression problems.

Parameters:

out is the output from the learned hypothesis.

y is the real output.

Returns:
Regression error between out and y. A commonly used measure is the squared error.

Reimplemented from LearnModel.
Definition at line 157 of file ordinal_ble.cpp.
References LearnModel::n_output(), and VALIDRANK.

void reset ( ) [virtual]

Cleaning up the learning model but keeping most settings.
Note:
This is probably needed after training or loading from file, but before having another training.

Reimplemented from LearnModel.
Definition at line 257 of file ordinal_ble.cpp.
References Ordinal_BLE::lm, and LearnModel::reset().

bool serialize ( std::ostream & ,

ver_list &

) const [protected, virtual]

Reimplemented from LearnModel.
Definition at line 18 of file ordinal_ble.cpp.
References Ordinal_BLE::full_ext, Ordinal_BLE::lm, Ordinal_BLE::n_ext, n_hyp, nrank, Ordinal_BLE::out_tab, and SERIALIZE_PARENT.

void set_full_extension ( bool = true )

Definition at line 105 of file ordinal_ble.cpp.
References Ordinal_BLE::ext_d, Ordinal_BLE::ext_w, and Ordinal_BLE::full_ext.

void set_model ( const LearnModel & )

set the underlying learning model

Definition at line 100 of file ordinal_ble.cpp.
References LearnModel::clone(), Ordinal_BLE::lm, and Ordinal_BLE::reset_data.

void set_tables ( BLE_TYPE ,

UINT

)

Definition at line 126 of file ordinal_ble.cpp.
References lemga::MULTI_THRESHOLD.

void set_tables ( const ECOC_TABLE & ,

const EXT_TABLE &

)

Definition at line 113 of file ordinal_ble.cpp.
References Ordinal_BLE::ext_tab, Ordinal_BLE::local_d, nrank, and Ordinal_BLE::out_tab.
Referenced by Ordinal_BLE::train().

void set_train_data ( const pDataSet & pd,

const pDataWgt & pw = 0

) [virtual]

Set the data set and sample weight to be used in training.
If the learning model/algorithm can only do training using uniform sample weight, i.e., support_weighted_data() returns false, a ``boostrapped'' copy of the original data set will be generated and used in the following training. The boostrapping is done by randomly pick samples (with replacement) w.r.t. the given weight pw.
In order to make the life easier, when support_weighted_data() returns true, a null pw will be replaced by a uniformly distributed probability vector. So we have the following invariant
Invariant:
support_weighted_data() == (ptw != 0)

Parameters:

pd gives the data set.

pw gives the sample weight, whose default value is 0.

See also:
support_weighted_data(), train()

Reimplemented from LearnModel.
Definition at line 162 of file ordinal_ble.cpp.
References Ordinal_BLE::d_nrank, Ordinal_BLE::ext_d, Ordinal_BLE::ext_w, LearnModel::n_samples, OUT2RANK, LearnModel::ptd, LearnModel::set_train_data(), and VALIDRANK.

virtual bool support_weighted_data ( ) const [inline, virtual]

Whether the learning model/algorithm supports unequally weighted data.

Returns:
true if supporting; false otherwise. The default is false, just for safety.

See also:
set_train_data()

Reimplemented from LearnModel.
Definition at line 59 of file ordinal_ble.h.

void train ( ) [virtual]

Train with preset data set and sample weight.

Implements LearnModel.
Definition at line 232 of file ordinal_ble.cpp.
References lemga::BLE_DEFAULT, Ordinal_BLE::d_nrank, Ordinal_BLE::ext_d, Ordinal_BLE::ext_w, Ordinal_BLE::extend_data(), Ordinal_BLE::lm, nrank, LearnModel::ptd, LearnModel::ptw, Ordinal_BLE::reset_data, LearnModel::set_dimensions(), and Ordinal_BLE::set_tables().

bool unserialize ( std::istream & ,

ver_list & ,

const id_t & = NIL_ID

) [protected, virtual]

Reimplemented from LearnModel.
Definition at line 45 of file ordinal_ble.cpp.
References Ordinal_BLE::d_nrank, Ordinal_BLE::ext_d, Ordinal_BLE::ext_tab, Ordinal_BLE::ext_w, Ordinal_BLE::full_ext, Ordinal_BLE::lm, Ordinal_BLE::n_ext, n_hyp, Object::NIL_ID, nrank, Ordinal_BLE::out_tab, LearnModel::ptd, LearnModel::ptw, Ordinal_BLE::reset_data, and UNSERIALIZE_PARENT.

Member Data Documentation

UINT d_nrank [protected]

number of ranking levels of the training set

Definition at line 70 of file ordinal_ble.h.
Referenced by Ordinal_BLE::n_rank(), Ordinal_BLE::operator=(), Ordinal_BLE::set_train_data(), Ordinal_BLE::train(), and Ordinal_BLE::unserialize().

pDataSet ext_d [protected]

the extended dataset

Definition at line 68 of file ordinal_ble.h.
Referenced by Ordinal_BLE::operator=(), Ordinal_BLE::set_full_extension(), Ordinal_BLE::set_train_data(), Ordinal_BLE::train(), and Ordinal_BLE::unserialize().

EXT_TABLE ext_tab [protected]

T (n_hyp) by E (n_ext) extension matrix.

Definition at line 28 of file ordinal_ble.h.
Referenced by Ordinal_BLE::extend_input(), Ordinal_BLE::extension_table(), Ordinal_BLE::operator=(), Ordinal_BLE::set_tables(), and Ordinal_BLE::unserialize().

pDataWgt ext_w [protected]

the weight for the extended set

Definition at line 69 of file ordinal_ble.h.
Referenced by Ordinal_BLE::operator=(), Ordinal_BLE::set_full_extension(), Ordinal_BLE::set_train_data(), Ordinal_BLE::train(), and Ordinal_BLE::unserialize().

bool full_ext [protected]

use the full extension or the partial one?

Definition at line 26 of file ordinal_ble.h.
Referenced by Ordinal_BLE::full_extension(), Ordinal_BLE::operator=(), Ordinal_BLE::serialize(), Ordinal_BLE::set_full_extension(), and Ordinal_BLE::unserialize().

pLearnModel lm [protected]

the learning model

Definition at line 25 of file ordinal_ble.h.
Referenced by Ordinal_BLE::distances(), Ordinal_BLE::model(), Ordinal_BLE::operator=(), Ordinal_BLE::Ordinal_BLE(), Ordinal_BLE::reset(), Ordinal_BLE::serialize(), Ordinal_BLE::set_model(), Ordinal_BLE::train(), and Ordinal_BLE::unserialize().

std::vector<REAL> local_d [mutable, protected]

Definition at line 78 of file ordinal_ble.h.
Referenced by Ordinal_BLE::set_tables().

UINT n_ext [protected]

Definition at line 29 of file ordinal_ble.h.
Referenced by Ordinal_BLE::operator=(), Ordinal_BLE::serialize(), and Ordinal_BLE::unserialize().

ECOC_TABLE out_tab [protected]

K (nrank) by T (n_hyp) output ECC matrix.

Definition at line 27 of file ordinal_ble.h.
Referenced by Ordinal_BLE::ECOC_table(), Ordinal_BLE::extend_example(), Ordinal_BLE::operator=(), Ordinal_BLE::serialize(), Ordinal_BLE::set_tables(), and Ordinal_BLE::unserialize().

bool reset_data [protected]

whether to reset the training set for lm

Definition at line 71 of file ordinal_ble.h.
Referenced by Ordinal_BLE::operator=(), Ordinal_BLE::set_model(), Ordinal_BLE::train(), and Ordinal_BLE::unserialize().

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

Generated on Wed Nov 8 08:17:19 2006 for LEMGA by

1.4.6

Ordinal_BLE Class Reference

Public Member Functions

Protected Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation