LTI-Lib latest version v1.9 - last update 10 Apr 2010

lti::combination Class Reference

This class is used to combine several results of different classifier::output objects. More...

#include <ltiCombination.h>

Inheritance diagram for lti::combination:
Inheritance graph
[legend]
Collaboration diagram for lti::combination:
Collaboration graph
[legend]

List of all members.

Classes

class  parameters
 the parameters for the class combination More...

Public Member Functions

 combination ()
 combination (const combination &other)
virtual ~combination ()
virtual const char * getTypeName () const
bool apply (const std::list< classifier::outputVector > &src, classifier::outputVector &dest) const
bool apply (const std::list< classifier::outputVector > &src, const dvector &weights, classifier::outputVector &dest) const
combinationcopy (const combination &other)
combinationoperator= (const combination &other)
virtual functorclone () const
const parametersgetParameters () const

Protected Member Functions

bool sumOfVectors (const std::list< classifier::outputVector > src, classifier::outputVector &dest) const
bool sumOfVectors (const std::list< classifier::outputVector > src, const dvector &weights, classifier::outputVector &dest) const
bool probBordaCount (const std::list< classifier::outputVector > src, classifier::outputVector &dest) const
bool probBordaCount (const std::list< classifier::outputVector > src, const dvector &weights, classifier::outputVector &dest) const
bool noMultAndCompatible (const std::list< classifier::outputVector > src) const

Detailed Description

This class is used to combine several results of different classifier::output objects.

There are many possibilities to do this. All of them will use internaly the classifier::sortedOutput type to compute the probability for each class. You can just combine linearly the results or use the probabilistic Borda count to consider also how sure is a classifier of its result.

Probabilistic Borda Count is documented in S. Syberichs. "Entscheidungs-Fusion zur 3D-Objekterkennung in einem hierarchischen Klassifikationssystem", Diplomarbeit, Lehrstuhl für Technische Informatik, RWTH-Aachen, Juni 2001


Constructor & Destructor Documentation

lti::combination::combination (  ) 

default constructor

lti::combination::combination ( const combination other  ) 

copy constructor

Parameters:
other the object to be copied
virtual lti::combination::~combination (  )  [virtual]

destructor


Member Function Documentation

bool lti::combination::apply ( const std::list< classifier::outputVector > &  src,
const dvector weights,
classifier::outputVector dest 
) const

Combine the given list of classifier::outputVector objects into one outputVector object according to the parameter settings.

No matter what the parameters are, the given weights will be used for combination.

Parameters:
src list of classifier::outputvector objects with the partial classification results.
weights used for weighting the inputs in same order.
dest classifier::outputvector where the result will be left.
Returns:
true if apply successful or false otherwise.
bool lti::combination::apply ( const std::list< classifier::outputVector > &  src,
classifier::outputVector dest 
) const

Combine the given list of classifier::outputVector objects into one outputVector object according to the parameter settings.

Parameters:
src list of classifier::outputvector objects with the partial classification results.
dest classifier::outputvector where the result will be left.
Returns:
true if apply successful or false otherwise.
virtual functor* lti::combination::clone (  )  const [virtual]

returns a pointer to a clone of this functor.

Implements lti::functor.

combination& lti::combination::copy ( const combination other  ) 

copy data of "other" functor.

Parameters:
other the functor to be copied
Returns:
a reference to this functor object

Reimplemented from lti::functor.

const parameters& lti::combination::getParameters (  )  const

returns used parameters

Reimplemented from lti::functor.

virtual const char* lti::combination::getTypeName (  )  const [virtual]

returns the name of this type ("combination")

Reimplemented from lti::functor.

bool lti::combination::noMultAndCompatible ( const std::list< classifier::outputVector src  )  const [protected]

Checks whether all outputVectors contained in src are compatible to each other and that none of them contains the same label more than once.

Returns true if everything is ok.

combination& lti::combination::operator= ( const combination other  ) 

alias for copy member

Parameters:
other the functor to be copied
Returns:
a reference to this functor object

Reimplemented from lti::functor.

bool lti::combination::probBordaCount ( const std::list< classifier::outputVector src,
const dvector weights,
classifier::outputVector dest 
) const [protected]

First calculates the ProbablilisticBordaCount for each input vector then sums these measures weighted by weights and returns the sum in dest.

Returns false if the vectors in src are not compatible, ie contain the same labels.

Parameters:
src list of outputVectors to be summed up
weights used for weighting the outputVectors in src
dest result of the sum
Returns:
false if values in list are incompatible
bool lti::combination::probBordaCount ( const std::list< classifier::outputVector src,
classifier::outputVector dest 
) const [protected]

First calculates the ProbablilisticBordaCount for each input vector then sums these measures and returns the sum in dest.

Returns false if the vectors in src are not compatible, ie contain the same labels.

Parameters:
src list of outputVectors to be summed up
dest result of the sum
Returns:
false if values in list are incompatible
bool lti::combination::sumOfVectors ( const std::list< classifier::outputVector src,
const dvector weights,
classifier::outputVector dest 
) const [protected]

Calculates the weighted sum of the outputVectors contained in src and leaves the result in dest.

Returns false if the vectors in src are not compatible, ie contain the same labels.

Parameters:
src list of outputVectors to be summed up
weights used for weighting the outputVectors in src
dest result of the sum
Returns:
false if values in list are incompatible
bool lti::combination::sumOfVectors ( const std::list< classifier::outputVector src,
classifier::outputVector dest 
) const [protected]

Calculates the sum of the outputVectors contained in src and leaves the result in dest.

Returns false if the vectors in src are not compatible, ie contain the same labels.

Parameters:
src list of outputVectors to be summed up
dest result of the sum
Returns:
false if values in list are incompatible

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

Generated on Sat Apr 10 15:26:47 2010 for LTI-Lib by Doxygen 1.6.1