moeoIndicatorBasedFitnessAssignment< MOEOT > Class Template Reference

Fitness assignment sheme based an Indicator proposed in: E. More...

#include <moeoIndicatorBasedFitnessAssignment.h>

Inheritance diagram for moeoIndicatorBasedFitnessAssignment< MOEOT >:


Public Types
typedef MOEOT::ObjectiveVector	ObjectiveVector
	The type of objective vector.
Public Member Functions
	moeoIndicatorBasedFitnessAssignment (moeoNormalizedSolutionVsSolutionBinaryMetric< ObjectiveVector, double > *_metric, const double _kappa)
	Ctor.
void	operator() (eoPop< MOEOT > &_pop)
	Sets the fitness values for every solution contained in the population _pop.
void	updateByDeleting (eoPop< MOEOT > &_pop, ObjectiveVector &_objVec)
	Updates the fitness values of the whole population _pop by taking the deletion of the objective vector _objVec into account.
double	updateByAdding (eoPop< MOEOT > &_pop, ObjectiveVector &_objVec)
	Updates the fitness values of the whole population _pop by taking the adding of the objective vector _objVec into account and returns the fitness value of _objVec.
Protected Member Functions
void	setup (const eoPop< MOEOT > &_pop)
	Sets the bounds for every objective using the min and the max value for every objective vector of _pop.
void	computeValues (const eoPop< MOEOT > &_pop)
	Compute every indicator value in values (values[i] = I(_v[i], _o)).
void	setFitnesses (eoPop< MOEOT > &_pop)
	Sets the fitness value of the whple population.
double	computeFitness (const unsigned _idx)
	Returns the fitness value of the _idx th individual of the population.
Protected Attributes
moeoNormalizedSolutionVsSolutionBinaryMetric< ObjectiveVector, double > *	metric
	the quality indicator
double	kappa
	the scaling factor
std::vector< std::vector< double > >	values
	the computed indicator values

Detailed Description

template<class MOEOT>
class moeoIndicatorBasedFitnessAssignment< MOEOT >

Fitness assignment sheme based an Indicator proposed in: E.

Zitzler, S. Künzli, "Indicator-Based Selection in Multiobjective Search", Proc. 8th International Conference on Parallel Problem Solving from Nature (PPSN VIII), pp. 832-842, Birmingham, UK (2004). This strategy is, for instance, used in IBEA.

Definition at line 28 of file moeoIndicatorBasedFitnessAssignment.h.

Constructor & Destructor Documentation

template<class MOEOT>

moeoIndicatorBasedFitnessAssignment< MOEOT >::moeoIndicatorBasedFitnessAssignment	(	moeoNormalizedSolutionVsSolutionBinaryMetric< ObjectiveVector, double > *	_metric,
		const double	_kappa
	)			`[inline]`

Ctor.

Parameters:

	_metric	the quality indicator
	_kappa	the scaling factor

Definition at line 41 of file moeoIndicatorBasedFitnessAssignment.h.

Member Function Documentation

template<class MOEOT>

void moeoIndicatorBasedFitnessAssignment< MOEOT >::operator() ( eoPop< MOEOT > & _pop ) [inline, virtual]

Sets the fitness values for every solution contained in the population _pop.

Parameters:

_pop

the population

Implements eoUF< eoPop< MOEOT > &, void >.

Definition at line 49 of file moeoIndicatorBasedFitnessAssignment.h.

References moeoIndicatorBasedFitnessAssignment< MOEOT >::computeValues(), moeoIndicatorBasedFitnessAssignment< MOEOT >::setFitnesses(), and moeoIndicatorBasedFitnessAssignment< MOEOT >::setup().

template<class MOEOT>

void moeoIndicatorBasedFitnessAssignment< MOEOT >::updateByDeleting	(	eoPop< MOEOT > &	_pop,
		ObjectiveVector &	_objVec
	)			`[inline, virtual]`

Updates the fitness values of the whole population _pop by taking the deletion of the objective vector _objVec into account.

Parameters:

	_pop	the population
	_objVec	the objective vector

Implements moeoFitnessAssignment< MOEOT >.

Definition at line 65 of file moeoIndicatorBasedFitnessAssignment.h.

References moeoIndicatorBasedFitnessAssignment< MOEOT >::kappa, and moeoIndicatorBasedFitnessAssignment< MOEOT >::metric.

template<class MOEOT>

double moeoIndicatorBasedFitnessAssignment< MOEOT >::updateByAdding	(	eoPop< MOEOT > &	_pop,
		ObjectiveVector &	_objVec
	)			`[inline]`

Updates the fitness values of the whole population _pop by taking the adding of the objective vector _objVec into account and returns the fitness value of _objVec.

Parameters: