MemeticAlgorithm< T > Class Template Reference

The super class for Memetic Algorithms. More...

#include <MemeticAlgorithm.h>

Inheritance diagram for MemeticAlgorithm< T >:

List of all members.

Public Types
enum	MA_STRATEGY
	Different strategy to select individuals undergoing local search. More...
Public Member Functions
	MemeticAlgorithm ()
	Default constructor.
	MemeticAlgorithm (GlobalSearch< T > *globalSearch, LocalSearch *localSearch=NULL)
	Constructor.
void	evolve (unsigned int nGenerations=1)
	Evolving algorithm.
int	nGenerations ()
	Number of generations elapsed.
bool	done ()
	Check if stopping criteria are satisfied.
Public Attributes
GlobalSearch< T > *	gs
	Global search method.
LocalSearch *	ls
	Local search method.
int	tLS
	Maximum number of evaluations allocated for each individual learning.
double	pLS
	Percentage of the population should undergo individual learning.
MA_STRATEGY	maStrategy
	Strategy to select individuals undergoing individual learning.
unsigned int	maxEvaluations
	(Stopping criteria) Maximum number of evaluations allowed.
unsigned int	maxGenerations
	(Stopping criteria) Maximum number of generations allowed.
Protected Member Functions
vector< int >	getLSIndividuals ()
	Generate the set of individuals undergoing local search, based on the setting of maStrategy.
virtual void	initialize ()
	Initialization.

---

Detailed Description

template<typename T>
class MemeticAlgorithm< T >

The super class for Memetic Algorithms.

<File comment goes here!!>

Copyright (c) 2008 by <Quang Huy / NTU>

Parameters:

T

Type T can be bool (binary problems), int (combinatorial problems) or real (continuous problems).

See also:

GeneticAlgorithm | LocalSearch

---

Member Enumeration Documentation

template<typename T>

enum MemeticAlgorithm::MA_STRATEGY

Different strategy to select individuals undergoing local search.

Best: individuals with best fitnesses are selected. Stratified: individuals are selected so as to maintain highest diversity. Random: individuals are selected randomly.

---

Constructor & Destructor Documentation

template<typename T>

MemeticAlgorithm< T >::MemeticAlgorithm	(	GlobalSearch< T > *	globalSearch,
		LocalSearch *	localSearch = NULL
	)			[inline]

Constructor.

Parameters:

	globalSearch	Global search method of type GlobalSearch.
	localSearch	Local search method of type LocalSearch.

Remarks:

Global search must be implemented for the same coding type.

See also:

GlobalSearch | LocalSearch

---

Member Function Documentation

template<typename T>

void MemeticAlgorithm< T >::evolve

(

unsigned int

nGenerations = 1

)

[inline]

Evolving algorithm.

Parameters:

nGenerations

Number of generation to evolve.

Typically, one of more generations of global search is followed by an individual learning phase.

Reimplemented in MAMetaLamarckian< T >.

template<typename T>

bool MemeticAlgorithm< T >::done

(

)

[inline]

Check if stopping criteria are satisfied.

Returns:

True if one of the stopping conditions is true, false otherwise.

template<typename T>

vector< int > MemeticAlgorithm< T >::getLSIndividuals

(

)

[inline, protected]

Generate the set of individuals undergoing local search, based on the setting of maStrategy.

Returns:

Vector of positions of the individuals.

See also:

MA_STRATEGY

---

Member Data Documentation

template<typename T>

GlobalSearch<T>* MemeticAlgorithm< T >::gs

Global search method.

See also:

GlobalSearch

template<typename T>

LocalSearch* MemeticAlgorithm< T >::ls

Local search method.

See also:

LocalSearch

template<typename T>

MA_STRATEGY MemeticAlgorithm< T >::maStrategy

Strategy to select individuals undergoing individual learning.

See also:

MA_STRATEGY

---

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

• D:/HUYNQ/Works/Codes/Dolphin/Dolphin/MemeticAlgorithms/MemeticAlgorithm.h