![]() |
Home | Main Page | Modules | Namespace List | Class Hierarchy | Alphabetical List | Data Structures | File List | Namespace Members | Data Fields | Globals | Related Pages |
#include <itkQuasiNewtonLBFGSOptimizer.h>
Static Public Member Functions | |
static Pointer | New () |
![]() | |
static Pointer | New () |
Protected Attributes | |
unsigned int | m_Bound |
DerivativeType | m_CurrentGradient |
unsigned long | m_CurrentIteration |
double | m_CurrentStepLength |
MeasureType | m_CurrentValue |
bool | m_InLineSearch |
unsigned int | m_Point |
unsigned int | m_PreviousPoint |
RhoType | m_Rho |
SType | m_S |
bool | m_Stop |
StopConditionType | m_StopCondition |
YType | m_Y |
![]() | |
ScaledCostFunctionPointer | m_ScaledCostFunction |
ParametersType | m_ScaledCurrentPosition |
Private Member Functions | |
void | operator= (const Self &) |
QuasiNewtonLBFGSOptimizer (const Self &) | |
Private Attributes | |
double | m_GradientMagnitudeTolerance |
LineSearchOptimizerPointer | m_LineSearchOptimizer |
unsigned long | m_MaximumNumberOfIterations |
unsigned int | m_Memory |
ITK version of the lbfgs algorithm ...
This class is an ITK version of the netlib lbfgs_ function. It gives exactly the same results, if used in combination with the itk::MoreThuenteLineSearchOptimizer.
The optimizer solves the unconstrained minimization problem
using the limited memory BFGS method. The routine is especially effective on problems involving a large number of variables. In a typical iteration of this method an approximation to the inverse of the Hessian is obtained by applying
BFGS updates to a diagonal matrix
, using information from the previous
steps. The user specifies the number
(Memory), which determines the amount of storage required by the routine.
The algorithm is described in "On the limited memory BFGS method for large scale optimization", by D. Liu and J. Nocedal, Mathematical Programming B 45 (1989) 503-528.
The steplength is determined at each iteration by means of a line search routine. The itk::MoreThuenteLineSearchOptimizer works well.
Definition at line 54 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef SmartPointer< const Self > itk::QuasiNewtonLBFGSOptimizer::ConstPointer |
Definition at line 61 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef Superclass::CostFunctionType itk::QuasiNewtonLBFGSOptimizer::CostFunctionType |
Definition at line 68 of file itkQuasiNewtonLBFGSOptimizer.h.
Definition at line 67 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef Array< double > itk::QuasiNewtonLBFGSOptimizer::DiagonalMatrixType |
Definition at line 76 of file itkQuasiNewtonLBFGSOptimizer.h.
Definition at line 79 of file itkQuasiNewtonLBFGSOptimizer.h.
Definition at line 77 of file itkQuasiNewtonLBFGSOptimizer.h.
Definition at line 70 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef Superclass::ParametersType itk::QuasiNewtonLBFGSOptimizer::ParametersType |
Definition at line 64 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef SmartPointer< Self > itk::QuasiNewtonLBFGSOptimizer::Pointer |
Definition at line 60 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef Array< double > itk::QuasiNewtonLBFGSOptimizer::RhoType |
Definition at line 73 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef Superclass::ScaledCostFunctionType itk::QuasiNewtonLBFGSOptimizer::ScaledCostFunctionType |
Definition at line 69 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef Superclass::ScalesType itk::QuasiNewtonLBFGSOptimizer::ScalesType |
Definition at line 71 of file itkQuasiNewtonLBFGSOptimizer.h.
Definition at line 58 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef std::vector< ParametersType > itk::QuasiNewtonLBFGSOptimizer::SType |
Definition at line 74 of file itkQuasiNewtonLBFGSOptimizer.h.
Definition at line 59 of file itkQuasiNewtonLBFGSOptimizer.h.
typedef std::vector< DerivativeType > itk::QuasiNewtonLBFGSOptimizer::YType |
Definition at line 75 of file itkQuasiNewtonLBFGSOptimizer.h.
Enumerator | |
---|---|
MetricError | |
LineSearchError | |
MaximumNumberOfIterations | |
InvalidDiagonalMatrix | |
GradientMagnitudeTolerance | |
ZeroStep | |
Unknown |
Definition at line 81 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
|
inlineprotectedvirtual |
Definition at line 131 of file itkQuasiNewtonLBFGSOptimizer.h.
|
private |
|
protectedvirtual |
Compute H0
Override this method if not satisfied with the default choice.
|
protectedvirtual |
Compute -Hg
COMPUTE -H*G USING THE FORMULA GIVEN IN: Nocedal, J. 1980, "Updating quasi-Newton matrices with limited storage", Mathematics of Computation, Vol.24, No.151, pp. 773-782.
|
virtual |
Reimplemented from itk::ScaledSingleValuedNonLinearOptimizer.
Reimplemented in elastix::QuasiNewtonLBFGS< TElastix >.
|
virtual |
|
virtual |
Get information about optimization process:
|
virtual |
|
virtual |
|
virtual |
Setting: the mininum gradient magnitude.
The optimizer stops when: ||CurrentGradient|| < GradientMagnitudeTolerance * max(1, ||CurrentPosition||)
|
virtual |
|
virtual |
|
virtual |
Setting: the maximum number of iterations
|
virtual |
|
virtual |
|
protectedvirtual |
Perform a line search along the search direction. On calling, x, f, and g should contain the current position, the cost function value at this position, and the derivative. On return the step, x (new position), f (value at x), and g (derivative at x) are updated.
Reimplemented in elastix::QuasiNewtonLBFGS< TElastix >.
|
static |
|
private |
|
inlineprotected |
Definition at line 134 of file itkQuasiNewtonLBFGSOptimizer.h.
|
virtual |
|
virtual |
|
protectedvirtual |
|
virtual |
Setting: the line search optimizer
|
virtual |
|
virtual |
Setting: the memory. The number of iterations that are used to estimate the Hessian. 5 by default. 0 results in (normalised) gradient descent search directions
|
virtual |
Reimplemented in elastix::QuasiNewtonLBFGS< TElastix >.
|
virtual |
|
protectedvirtual |
Store s = x_k - x_k-1 and y = g_k - g_k-1 in m_S and m_Y, and store 1/(ys) in m_Rho.
|
protectedvirtual |
Check if convergence has occured; The firstLineSearchDone bool allows the implementation of TestConvergence to decide to skip a few convergence checks when no line search has performed yet (so, before the actual optimisation begins)
Reimplemented in elastix::QuasiNewtonLBFGS< TElastix >.
|
protected |
Definition at line 152 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 136 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 138 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 141 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 137 of file itkQuasiNewtonLBFGSOptimizer.h.
|
private |
Definition at line 201 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Is true when the LineSearchOptimizer has been started.
Definition at line 144 of file itkQuasiNewtonLBFGSOptimizer.h.
|
private |
Definition at line 202 of file itkQuasiNewtonLBFGSOptimizer.h.
|
private |
Definition at line 200 of file itkQuasiNewtonLBFGSOptimizer.h.
|
private |
Definition at line 203 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 150 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 151 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 146 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 147 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 140 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 139 of file itkQuasiNewtonLBFGSOptimizer.h.
|
protected |
Definition at line 148 of file itkQuasiNewtonLBFGSOptimizer.h.
Generated on 27-04-2014 for elastix by ![]() |
![]() |