#include <itkInitializationBiasedParticleSwarmOptimizer.h>

Inheritance diagram for itk::InitializationBiasedParticleSwarmOptimizer:

Collaboration diagram for itk::InitializationBiasedParticleSwarmOptimizer:

Detailed Description

Implementation of a biased/regularized Particle Swarm Optimization (PSO) algorithm.

This PSO algorithm was originally described in: M. P. Wachowiak, R. Smolikova, Y. Zheng, J. M. Zurada, A. S. Elmaghraby, "An approach to multimodal biomedical image registration utilizing particle swarm optimization", IEEE Trans. Evol. Comput., vol. 8(3): 289-301, 2004.

The algorithm uses a stochastic optimization approach. Optimization is performed by maintaining a swarm (flock) of particles that traverse the parameter space, searching for the optimal function value. Associated with each particle are its location and speed, in parameter space. A particle's next location is determined by its current location, its current speed, the location of the best function value it previously encountered, the location of the best function value the particles in its neighborhood previously encountered and the initial position the user specified.

The assumption is that the user's initial parameter settings are close to the minimum, which is often the case for registration. The initial parameter values are incorporated into the PSO's update rules, biasing the search in their direction. The swarms update equations are thus:

$v_i(t+1) = wv_i(t) + c_1u_1(p_i-x_i(t)) + c_2u_2(p_g-x_i(t)) + c_3u_3(x_{init} - x_i(t))$ $x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $u_i$ are $~U(0,1)$ and $w,c_1,c_2, c_3$ are user selected weights, and c_3 is linearly decreased per iteration so that it is in $c_3=initial, 0$ .

Swarm initialization is performed within the user supplied parameter bounds using a uniform distribution or a normal distribution centered on the initial parameter values supplied by the user, $x_{init}$ . The search terminates when the maximal number of iterations has been reached or when the change in the best value in the past $g$ generations is below a threshold and the swarm has collapsed (i.e. particles are close to each other in parameter space).

Note: This implementation only performs minimization.

Definition at line 71 of file itkInitializationBiasedParticleSwarmOptimizer.h.

Public Types
typedef double	CoefficientType

typedef SmartPointer< const Self >	ConstPointer

typedef SmartPointer< Self >	Pointer

typedef InitializationBiasedParticleSwarmOptimizer	Self

typedef ParticleSwarmOptimizerBase	Superclass

Public Types inherited from itk::ParticleSwarmOptimizerBase
typedef SmartPointer< const Self >	ConstPointer

typedef CostFunctionType::MeasureType	MeasureType

typedef unsigned int	NumberOfGenerationsType

typedef unsigned int	NumberOfIterationsType

typedef unsigned int	NumberOfParticlesType

typedef std::vector< std::pair < ParametersType::ValueType, ParametersType::ValueType > >	ParameterBoundsType

typedef SmartPointer< Self >	Pointer

typedef Statistics::MersenneTwisterRandomVariateGenerator	RandomVariateGeneratorType

typedef ParticleSwarmOptimizerBase	Self

typedef SingleValuedNonLinearOptimizer	Superclass

typedef std::vector< ParticleData >	SwarmType

typedef ParametersType::ValueType	ValueType

Public Types inherited from itk::SingleValuedNonLinearOptimizer
typedef SmartPointer< const Self >	ConstPointer

typedef CostFunctionType::Pointer	CostFunctionPointer

typedef SingleValuedCostFunction	CostFunctionType

typedef CostFunctionType::DerivativeType	DerivativeType

typedef CostFunctionType::MeasureType	MeasureType

typedef Superclass::ParametersType	ParametersType

typedef SmartPointer< Self >	Pointer

typedef SingleValuedNonLinearOptimizer	Self

typedef NonLinearOptimizer	Superclass

Public Types inherited from itk::NonLinearOptimizer
typedef SmartPointer< const Self >	ConstPointer

typedef Superclass::ParametersType	ParametersType

typedef SmartPointer< Self >	Pointer

typedef Superclass::ScalesType	ScalesType

typedef NonLinearOptimizer	Self

typedef Optimizer	Superclass

Public Types inherited from itk::Optimizer
typedef SmartPointer< const Self >	ConstPointer

typedef OptimizerParameters < double >	ParametersType

typedef SmartPointer< Self >	Pointer

typedef Array< double >	ScalesType

typedef Optimizer	Self

typedef Object	Superclass

Public Types inherited from itk::Object
typedef SmartPointer< const Self >	ConstPointer

typedef SmartPointer< Self >	Pointer

typedef Object	Self

typedef LightObject	Superclass

Public Types inherited from itk::LightObject
typedef SmartPointer< const Self >	ConstPointer

typedef SmartPointer< Self >	Pointer

typedef LightObject	Self

Public Member Functions
virtual ::itk::LightObject::Pointer	CreateAnother (void) const

virtual const char *	GetNameOfClass () const


virtual void	SetInertiaCoefficient (CoefficientType _arg)

virtual CoefficientType	GetInertiaCoefficient ()

virtual void	SetPersonalCoefficient (CoefficientType _arg)

virtual CoefficientType	GetPersonalCoefficient ()

virtual void	SetGlobalCoefficient (CoefficientType _arg)

virtual CoefficientType	GetGlobalCoefficient ()

virtual void	SetInitializationCoefficient (CoefficientType _arg)

virtual CoefficientType	GetInitializationCoefficient ()

Public Member Functions inherited from itk::ParticleSwarmOptimizerBase
virtual MeasureType	GetFunctionConvergenceTolerance ()

virtual bool	GetInitializeNormalDistribution ()

virtual NumberOfIterationsType	GetMaximalNumberOfIterations ()

virtual NumberOfGenerationsType	GetNumberOfGenerationsWithMinimalImprovement ()

virtual NumberOfParticlesType	GetNumberOfParticles ()

ParameterBoundsType	GetParameterBounds () const

virtual ParametersType	GetParametersConvergenceTolerance ()

virtual double	GetPercentageParticlesConverged ()

virtual bool	GetPrintSwarm ()

virtual RandomVariateGeneratorType::IntegerType	GetSeed ()

virtual const std::string	GetStopConditionDescription () const

virtual bool	GetUseSeed ()

MeasureType	GetValue () const

virtual void	InitializeNormalDistributionOff ()

virtual void	InitializeNormalDistributionOn ()

void	PrintSwarm (std::ostream &os, Indent indent) const

virtual void	PrintSwarmOff ()

virtual void	PrintSwarmOn ()

virtual void	SetFunctionConvergenceTolerance (MeasureType _arg)

virtual void	SetInitializeNormalDistribution (bool _arg)

virtual void	SetMaximalNumberOfIterations (NumberOfIterationsType _arg)

virtual void	SetNumberOfGenerationsWithMinimalImprovement (NumberOfGenerationsType _arg)

void	SetNumberOfParticles (NumberOfParticlesType n)

void	SetParametersConvergenceTolerance (ValueType convergenceTolerance, unsigned int sz)

virtual void	SetParametersConvergenceTolerance (ParametersType _arg)

virtual void	SetPercentageParticlesConverged (double _arg)

virtual void	SetPrintSwarm (bool _arg)

virtual void	SetSeed (RandomVariateGeneratorType::IntegerType _arg)

virtual void	SetUseSeed (bool _arg)

void	StartOptimization (void)

virtual void	UseSeedOff ()

virtual void	UseSeedOn ()

void	SetInitialSwarm (const SwarmType &initialSwarm)

void	ClearSwarm ()

virtual void	SetParameterBounds (ParameterBoundsType &bounds)

void	SetParameterBounds (std::pair< ParametersType::ValueType, ParametersType::ValueType > &bounds, unsigned int n)

Public Member Functions inherited from itk::SingleValuedNonLinearOptimizer
virtual ::itk::LightObject::Pointer	CreateAnother (void) const

virtual const CostFunctionType *	GetCostFunction () const

virtual CostFunctionType *	GetModifiableCostFunction ()

MeasureType	GetValue (const ParametersType &parameters) const

virtual void	SetCostFunction (CostFunctionType *costFunction)

Static Public Member Functions
static Pointer	New ()

Protected Member Functions
	InitializationBiasedParticleSwarmOptimizer ()

void	PrintSelf (std::ostream &os, Indent indent) const

virtual void	UpdateSwarm ()

virtual	~InitializationBiasedParticleSwarmOptimizer ()

Protected Member Functions inherited from itk::ParticleSwarmOptimizerBase
void	FileInitialization ()

virtual void	Initialize ()

void	operator= (const Self &)

	ParticleSwarmOptimizerBase ()

	ParticleSwarmOptimizerBase (const Self &)

void	PrintParamtersType (const ParametersType &x, std::ostream &os) const

void	PrintSelf (std::ostream &os, Indent indent) const

void	RandomInitialization ()

virtual void	ValidateSettings ()

virtual	~ParticleSwarmOptimizerBase ()

Protected Member Functions inherited from itk::SingleValuedNonLinearOptimizer
void	PrintSelf (std::ostream &os, Indent indent) const

	SingleValuedNonLinearOptimizer ()

virtual	~SingleValuedNonLinearOptimizer ()

Protected Member Functions inherited from itk::NonLinearOptimizer
	NonLinearOptimizer ()

virtual	~NonLinearOptimizer ()

Protected Member Functions inherited from itk::Optimizer
	Optimizer ()

virtual void	SetCurrentPosition (const ParametersType &param)

virtual	~Optimizer ()

Protected Member Functions inherited from itk::Object
	Object ()

bool	PrintObservers (std::ostream &os, Indent indent) const

virtual void	SetTimeStamp (const TimeStamp &time)

virtual	~Object ()

Protected Member Functions inherited from itk::LightObject
virtual LightObject::Pointer	InternalClone () const

	LightObject ()

virtual void	PrintHeader (std::ostream &os, Indent indent) const

virtual void	PrintTrailer (std::ostream &os, Indent indent) const

virtual	~LightObject ()

Private Member Functions
	InitializationBiasedParticleSwarmOptimizer (const Self &)

void	operator= (const Self &)

Private Attributes
ParametersType::ValueType	m_GlobalCoefficient

ParametersType::ValueType	m_InertiaCoefficient

ParametersType::ValueType	m_InitializationCoefficient

ParametersType::ValueType	m_PersonalCoefficient

Additional Inherited Members
Protected Types inherited from itk::LightObject
typedef int	InternalReferenceCountType

Protected Attributes inherited from itk::ParticleSwarmOptimizerBase
CostFunctionType::MeasureType	m_FunctionBestValue

std::vector< MeasureType >	m_FunctionBestValueMemory

CostFunctionType::MeasureType	m_FunctionConvergenceTolerance

bool	m_InitializeNormalDistribution

NumberOfIterationsType	m_IterationIndex

NumberOfIterationsType	m_MaximalNumberOfIterations

NumberOfGenerationsType	m_NumberOfGenerationsWithMinimalImprovement

NumberOfParticlesType	m_NumberOfParticles

ParameterBoundsType	m_ParameterBounds

ParametersType	m_ParametersBestValue

ParametersType	m_ParametersConvergenceTolerance

std::vector< ParticleData >	m_Particles

double	m_PercentageParticlesConverged

bool	m_PrintSwarm

RandomVariateGeneratorType::IntegerType	m_Seed

std::ostringstream	m_StopConditionDescription

bool	m_UseSeed

Member Typedef Documentation

typedef double itk::InitializationBiasedParticleSwarmOptimizer::CoefficientType

Definition at line 81 of file itkInitializationBiasedParticleSwarmOptimizer.h.

typedef SmartPointer<const Self> itk::InitializationBiasedParticleSwarmOptimizer::ConstPointer

Definition at line 79 of file itkInitializationBiasedParticleSwarmOptimizer.h.

typedef SmartPointer<Self> itk::InitializationBiasedParticleSwarmOptimizer::Pointer

Definition at line 78 of file itkInitializationBiasedParticleSwarmOptimizer.h.

typedef InitializationBiasedParticleSwarmOptimizer itk::InitializationBiasedParticleSwarmOptimizer::Self

Standard "Self" typedef.

Definition at line 76 of file itkInitializationBiasedParticleSwarmOptimizer.h.

typedef ParticleSwarmOptimizerBase itk::InitializationBiasedParticleSwarmOptimizer::Superclass

Definition at line 77 of file itkInitializationBiasedParticleSwarmOptimizer.h.

Constructor & Destructor Documentation

itk::InitializationBiasedParticleSwarmOptimizer::InitializationBiasedParticleSwarmOptimizer ( )

protected

virtual itk::InitializationBiasedParticleSwarmOptimizer::~InitializationBiasedParticleSwarmOptimizer ( )

inlineprotectedvirtual

Definition at line 118 of file itkInitializationBiasedParticleSwarmOptimizer.h.

itk::InitializationBiasedParticleSwarmOptimizer::InitializationBiasedParticleSwarmOptimizer ( const Self & )

private

Member Function Documentation

virtual::itk::LightObject::Pointer itk::InitializationBiasedParticleSwarmOptimizer::CreateAnother ( void ) const

virtual

Create an object from an instance, potentially deferring to a factory. This method allows you to create an instance of an object that is exactly the same type as the referring object. This is useful in cases where an object has been cast back to a base class.

Reimplemented from itk::Object.

virtual CoefficientType itk::InitializationBiasedParticleSwarmOptimizer::GetGlobalCoefficient ( )

virtual

The Particle swarm optimizer uses the following update formula:

$c_3 = c_{3initial}(1.0 - IterationIndex/MaximalNumberOfIterations)$

$v_i(t+1) = w*v_i(t) + c_1*uniform(0,1)*(p_i-x_i(t)) + c_2*uniform(0,1)*(p_g-x_i(t)) + c_3*uniform(0,1)*(x_{init}-x_i(t))$

$x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $w$ - inertia constant $c_1$ - personal coefficient $c_2$ - global coefficient $c_3$ - initial location coefficient $p_i$ - parameters yielding the best function value obtained by this particle $p_g$ - parameters yielding the best function value obtained by all particles $x_{init}$ - initial parameter values provided by user

virtual CoefficientType itk::InitializationBiasedParticleSwarmOptimizer::GetInertiaCoefficient ( )

virtual

The Particle swarm optimizer uses the following update formula:

$c_3 = c_{3initial}(1.0 - IterationIndex/MaximalNumberOfIterations)$

$v_i(t+1) = w*v_i(t) + c_1*uniform(0,1)*(p_i-x_i(t)) + c_2*uniform(0,1)*(p_g-x_i(t)) + c_3*uniform(0,1)*(x_{init}-x_i(t))$

$x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $w$ - inertia constant $c_1$ - personal coefficient $c_2$ - global coefficient $c_3$ - initial location coefficient $p_i$ - parameters yielding the best function value obtained by this particle $p_g$ - parameters yielding the best function value obtained by all particles $x_{init}$ - initial parameter values provided by user

virtual CoefficientType itk::InitializationBiasedParticleSwarmOptimizer::GetInitializationCoefficient ( )

virtual

The Particle swarm optimizer uses the following update formula:

$c_3 = c_{3initial}(1.0 - IterationIndex/MaximalNumberOfIterations)$

$v_i(t+1) = w*v_i(t) + c_1*uniform(0,1)*(p_i-x_i(t)) + c_2*uniform(0,1)*(p_g-x_i(t)) + c_3*uniform(0,1)*(x_{init}-x_i(t))$

$x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $w$ - inertia constant $c_1$ - personal coefficient $c_2$ - global coefficient $c_3$ - initial location coefficient $p_i$ - parameters yielding the best function value obtained by this particle $p_g$ - parameters yielding the best function value obtained by all particles $x_{init}$ - initial parameter values provided by user

virtual const char* itk::InitializationBiasedParticleSwarmOptimizer::GetNameOfClass ( ) const

virtual

Run-time type information (and related methods).

Reimplemented from itk::ParticleSwarmOptimizerBase.

virtual CoefficientType itk::InitializationBiasedParticleSwarmOptimizer::GetPersonalCoefficient ( )

virtual

The Particle swarm optimizer uses the following update formula:

$c_3 = c_{3initial}(1.0 - IterationIndex/MaximalNumberOfIterations)$

$v_i(t+1) = w*v_i(t) + c_1*uniform(0,1)*(p_i-x_i(t)) + c_2*uniform(0,1)*(p_g-x_i(t)) + c_3*uniform(0,1)*(x_{init}-x_i(t))$

$x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $w$ - inertia constant $c_1$ - personal coefficient $c_2$ - global coefficient $c_3$ - initial location coefficient $p_i$ - parameters yielding the best function value obtained by this particle $p_g$ - parameters yielding the best function value obtained by all particles $x_{init}$ - initial parameter values provided by user

static Pointer itk::InitializationBiasedParticleSwarmOptimizer::New ( )

static

Method for creation through the object factory.

void itk::InitializationBiasedParticleSwarmOptimizer::operator= ( const Self & )

private

void itk::InitializationBiasedParticleSwarmOptimizer::PrintSelf	(	std::ostream &	os,
		Indent	indent
	)		const

protectedvirtual

Methods invoked by Print() to print information about the object including superclasses. Typically not called by the user (use Print() instead) but used in the hierarchical print process to combine the output of several classes.

Reimplemented from itk::Object.

virtual void itk::InitializationBiasedParticleSwarmOptimizer::SetGlobalCoefficient ( CoefficientType _arg )

virtual

The Particle swarm optimizer uses the following update formula:

$c_3 = c_{3initial}(1.0 - IterationIndex/MaximalNumberOfIterations)$

$v_i(t+1) = w*v_i(t) + c_1*uniform(0,1)*(p_i-x_i(t)) + c_2*uniform(0,1)*(p_g-x_i(t)) + c_3*uniform(0,1)*(x_{init}-x_i(t))$

$x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $w$ - inertia constant $c_1$ - personal coefficient $c_2$ - global coefficient $c_3$ - initial location coefficient $p_i$ - parameters yielding the best function value obtained by this particle $p_g$ - parameters yielding the best function value obtained by all particles $x_{init}$ - initial parameter values provided by user

virtual void itk::InitializationBiasedParticleSwarmOptimizer::SetInertiaCoefficient ( CoefficientType _arg )

virtual

The Particle swarm optimizer uses the following update formula:

$c_3 = c_{3initial}(1.0 - IterationIndex/MaximalNumberOfIterations)$

$v_i(t+1) = w*v_i(t) + c_1*uniform(0,1)*(p_i-x_i(t)) + c_2*uniform(0,1)*(p_g-x_i(t)) + c_3*uniform(0,1)*(x_{init}-x_i(t))$

$x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $w$ - inertia constant $c_1$ - personal coefficient $c_2$ - global coefficient $c_3$ - initial location coefficient $p_i$ - parameters yielding the best function value obtained by this particle $p_g$ - parameters yielding the best function value obtained by all particles $x_{init}$ - initial parameter values provided by user

virtual void itk::InitializationBiasedParticleSwarmOptimizer::SetInitializationCoefficient ( CoefficientType _arg )

virtual

The Particle swarm optimizer uses the following update formula:

$c_3 = c_{3initial}(1.0 - IterationIndex/MaximalNumberOfIterations)$

$v_i(t+1) = w*v_i(t) + c_1*uniform(0,1)*(p_i-x_i(t)) + c_2*uniform(0,1)*(p_g-x_i(t)) + c_3*uniform(0,1)*(x_{init}-x_i(t))$

$x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $w$ - inertia constant $c_1$ - personal coefficient $c_2$ - global coefficient $c_3$ - initial location coefficient $p_i$ - parameters yielding the best function value obtained by this particle $p_g$ - parameters yielding the best function value obtained by all particles $x_{init}$ - initial parameter values provided by user

virtual void itk::InitializationBiasedParticleSwarmOptimizer::SetPersonalCoefficient ( CoefficientType _arg )

virtual

The Particle swarm optimizer uses the following update formula:

$c_3 = c_{3initial}(1.0 - IterationIndex/MaximalNumberOfIterations)$

$v_i(t+1) = w*v_i(t) + c_1*uniform(0,1)*(p_i-x_i(t)) + c_2*uniform(0,1)*(p_g-x_i(t)) + c_3*uniform(0,1)*(x_{init}-x_i(t))$

$x_i(t+1) = clampToBounds(x_i(t) + v_i(t+1))$

where $w$ - inertia constant $c_1$ - personal coefficient $c_2$ - global coefficient $c_3$ - initial location coefficient $p_i$ - parameters yielding the best function value obtained by this particle $p_g$ - parameters yielding the best function value obtained by all particles $x_{init}$ - initial parameter values provided by user

virtual void itk::InitializationBiasedParticleSwarmOptimizer::UpdateSwarm ( )

protectedvirtual

Implement your update rule in this function.

Implements itk::ParticleSwarmOptimizerBase.