GClasses::GMomentumGreedySearch Class Reference

At each iteration this algorithm moves in only one dimension. If the situation doesn't improve it tries the opposite direction. If both directions are worse, it decreases the step size for that dimension, otherwise it increases the step size for that dimension. More...

#include <GHillClimber.h>

Inheritance diagram for GClasses::GMomentumGreedySearch:

Public Member Functions
	GMomentumGreedySearch (GTargetFunction *pCritic)
virtual	~GMomentumGreedySearch ()
virtual double *	currentVector ()
	Returns a pointer to the state vector. More...
virtual double	iterate ()
	Makes another attempt to find a better vector. Returns the heuristic error. (Usually you will call this method in a loop until your stopping criteria has been met.) More...
void	setAllStepSizes (double dStepSize)
	Set all the current step sizes to this value. More...
void	setChangeFactor (double d)
	d should be a value between 0 and 1 More...
double *	stepSizes ()
	Returns the vector of step sizes. More...
Public Member Functions inherited from GClasses::GOptimizer
	GOptimizer (GTargetFunction *pCritic)
virtual	~GOptimizer ()
double	searchUntil (size_t nBurnInIterations, size_t nIterations, double dImprovement)
	This will first call iterate() nBurnInIterations times, then it will repeatedly call iterate() in blocks of nIterations times. If the error heuristic has not improved by the specified ratio after a block of iterations, it will stop. (For example, if the error before the block of iterations was 50, and the error after is 49, then training will stop if dImprovement is > 0.02.) If the error heuristic is not stable, then the value of nIterations should be large. More...

Protected Member Functions
double	iterateOneDim ()
void	reset ()

Protected Attributes
double	m_dChangeFactor
double	m_dError
size_t	m_nCurrentDim
size_t	m_nDimensions
double *	m_pStepSizes
double *	m_pVector
Protected Attributes inherited from GClasses::GOptimizer
GTargetFunction *	m_pCritic

Detailed Description

At each iteration this algorithm moves in only one dimension. If the situation doesn't improve it tries the opposite direction. If both directions are worse, it decreases the step size for that dimension, otherwise it increases the step size for that dimension.

Constructor & Destructor Documentation

GClasses::GMomentumGreedySearch::GMomentumGreedySearch

(

GTargetFunction *

pCritic

)

virtual GClasses::GMomentumGreedySearch::~GMomentumGreedySearch ( )

virtual

Member Function Documentation

virtual double* GClasses::GMomentumGreedySearch::currentVector ( )

inlinevirtual

Returns a pointer to the state vector.

Implements GClasses::GOptimizer.

virtual double GClasses::GMomentumGreedySearch::iterate ( )

virtual

Makes another attempt to find a better vector. Returns the heuristic error. (Usually you will call this method in a loop until your stopping criteria has been met.)

Implements GClasses::GOptimizer.

double GClasses::GMomentumGreedySearch::iterateOneDim ( )

protected

void GClasses::GMomentumGreedySearch::reset ( )

protected

void GClasses::GMomentumGreedySearch::setAllStepSizes

(

double

dStepSize

)

Set all the current step sizes to this value.

void GClasses::GMomentumGreedySearch::setChangeFactor ( double  d )

inline

d should be a value between 0 and 1

double* GClasses::GMomentumGreedySearch::stepSizes

(

)

Returns the vector of step sizes.

Member Data Documentation

double GClasses::GMomentumGreedySearch::m_dChangeFactor

protected

double GClasses::GMomentumGreedySearch::m_dError

protected

size_t GClasses::GMomentumGreedySearch::m_nCurrentDim

protected

size_t GClasses::GMomentumGreedySearch::m_nDimensions

protected

double* GClasses::GMomentumGreedySearch::m_pStepSizes

protected

double* GClasses::GMomentumGreedySearch::m_pVector

protected