shark::QpSolver< Problem, SelectionStrategy > Class Template Reference

Quadratic program solver. More...

#include <shark/Algorithms/QP/QpSolver.h>

Public Member Functions
	QpSolver (Problem &problem)
void	solve (QpStoppingCondition &stop, QpSolutionProperties *prop=NULL)
	Solve the quadratic program. More...

Protected Attributes
Problem &	m_problem

Detailed Description

template<class Problem, class SelectionStrategy = typename Problem::PreferedSelectionStrategy>
class shark::QpSolver< Problem, SelectionStrategy >

Quadratic program solver.

todo: new documentation

Definition at line 403 of file QpSolver.h.

Constructor & Destructor Documentation

QpSolver()

template<class Problem, class SelectionStrategy = typename Problem::PreferedSelectionStrategy>

shark::QpSolver< Problem, SelectionStrategy >::QpSolver ( Problem &  problem )

inline

Definition at line 406 of file QpSolver.h.

Member Function Documentation

solve()

template<class Problem, class SelectionStrategy = typename Problem::PreferedSelectionStrategy>

void shark::QpSolver< Problem, SelectionStrategy >::solve ( QpStoppingCondition &  stop, QpSolutionProperties *  prop = NULL  )

inline

Solve the quadratic program.

The function iteratively refines the solution by applying the SMO (subset descent) algorithm until one of the stopping criteria is fulfilled.

Definition at line 415 of file QpSolver.h.

References shark::QpStoppingCondition::maxIterations, shark::QpStoppingCondition::maxSeconds, shark::QpStoppingCondition::minAccuracy, shark::Timer::now(), shark::QpAccuracyReached, shark::QpMaxIterationsReached, and shark::QpTimeout.

Referenced by shark::RadiusMarginQuotient< InputType, CacheType >::computeRadiusMargin(), shark::CSvmTrainer< InputType, CacheType >::get_db_dParams(), shark::MissingFeatureSvmTrainer< InputType, CacheType >::setMaxIterations(), shark::BiasSolver< Matrix >::solve(), shark::BiasSolverSimplex< Matrix >::solve(), shark::EpsilonSvmTrainer< InputType, CacheType >::train(), shark::RankingSvmTrainer< InputType, CacheType >::train(), and shark::SquaredHingeCSvmTrainer< InputType, CacheType >::train().

Member Data Documentation

m_problem

template<class Problem, class SelectionStrategy = typename Problem::PreferedSelectionStrategy>

Problem& shark::QpSolver< Problem, SelectionStrategy >::m_problem

protected

Definition at line 481 of file QpSolver.h.

---

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

• include/shark/Algorithms/QP/QpSolver.h