mlpack::fastmks::FastMKSRules< KernelType, TreeType > Class Template Reference

The FastMKSRules class is a template helper class used by FastMKS class when performing exact max-kernel search. More...

#include <fastmks_rules.hpp>

Public Types
typedef tree::TraversalInfo< TreeType >	TraversalInfoType

Public Member Functions
	FastMKSRules (const typename TreeType::Mat &referenceSet, const typename TreeType::Mat &querySet, const size_t k, KernelType &kernel)
	Construct the FastMKSRules object. More...
void	GetResults (arma::Mat< size_t > &indices, arma::mat &products)
	Store the list of candidates for each query point in the given matrices. More...
double	BaseCase (const size_t queryIndex, const size_t referenceIndex)
	Compute the base case (kernel value) between two points.
double	Score (const size_t queryIndex, TreeType &referenceNode)
	Get the score for recursion order. More...
double	Score (TreeType &queryNode, TreeType &referenceNode)
	Get the score for recursion order. More...
double	Rescore (const size_t queryIndex, TreeType &referenceNode, const double oldScore) const
	Re-evaluate the score for recursion order. More...
double	Rescore (TreeType &queryNode, TreeType &referenceNode, const double oldScore) const
	Re-evaluate the score for recursion order. More...
size_t	BaseCases () const
	Get the number of times BaseCase() was called.
size_t &	BaseCases ()
	Modify the number of times BaseCase() was called.
size_t	Scores () const
	Get the number of times Score() was called.
size_t &	Scores ()
	Modify the number of times Score() was called.
const TraversalInfoType &	TraversalInfo () const
TraversalInfoType &	TraversalInfo ()
size_t	MinimumBaseCases () const
	Get the minimum number of base cases we need to perform to have acceptable results. More...

Detailed Description

template<typename KernelType, typename TreeType>
class mlpack::fastmks::FastMKSRules< KernelType, TreeType >

The FastMKSRules class is a template helper class used by FastMKS class when performing exact max-kernel search.

For each point in the query dataset, it keeps track of the k best candidates in the reference dataset.

Template Parameters

KernelType	Type of kernel to run FastMKS with.
TreeType	Type of tree to run FastMKS with; it must satisfy the TreeType policy API.

Constructor & Destructor Documentation

FastMKSRules()

template<typename KernelType , typename TreeType >

mlpack::fastmks::FastMKSRules< KernelType, TreeType >::FastMKSRules	(	const typename TreeType::Mat &	referenceSet,
		const typename TreeType::Mat &	querySet,
		const size_t	k,
		KernelType &	kernel
	)

Construct the FastMKSRules object.

This is usually done from within the FastMKS class at search time.

Parameters

referenceSet	Set of reference data.
querySet	Set of query data.
k	Number of candidates to search for.
kernel	Kernel to run FastMKS with.

Member Function Documentation

GetResults()

template<typename KernelType , typename TreeType >

void mlpack::fastmks::FastMKSRules< KernelType, TreeType >::GetResults	(	arma::Mat< size_t > &	indices,
		arma::mat &	products
	)

Store the list of candidates for each query point in the given matrices.

Parameters

indices	Matrix storing lists of candidate for each query point.
products	Matrix storing kernel value for each candidate.

MinimumBaseCases()

template<typename KernelType , typename TreeType >

size_t mlpack::fastmks::FastMKSRules< KernelType, TreeType >::MinimumBaseCases ( ) const

inline

Get the minimum number of base cases we need to perform to have acceptable results.

Rescore() [1/2]

template<typename KernelType , typename TreeType >

double mlpack::fastmks::FastMKSRules< KernelType, TreeType >::Rescore	(	const size_t	queryIndex,
		TreeType &	referenceNode,
		const double	oldScore
	)		const

Re-evaluate the score for recursion order.

A low score indicates priority for recursion, while DBL_MAX indicates that a node should not be recursed into at all (it should be pruned). This is used when the score has already been calculated, but another recursion may have modified the bounds for pruning. So the old score is checked against the new pruning bound.

Parameters

queryIndex	Index of query point.
referenceNode	Candidate node to be recursed into.
oldScore	Old score produced by Score() (or Rescore()).

Rescore() [2/2]

template<typename KernelType , typename TreeType >

double mlpack::fastmks::FastMKSRules< KernelType, TreeType >::Rescore	(	TreeType &	queryNode,
		TreeType &	referenceNode,
		const double	oldScore
	)		const

Re-evaluate the score for recursion order.

A low score indicates priority for recursion, while DBL_MAX indicates that a node should not be recursed into at all (it should be pruned). This is used when the score has already been calculated, but another recursion may have modified the bounds for pruning. So the old score is checked against the new pruning bound.

Parameters

queryNode	Candidate query node to be recursed into.
referenceNode	Candidate reference node to be recursed into.
oldScore	Old score produced by Score() (or Rescore()).

Score() [1/2]

template<typename KernelType , typename TreeType >

double mlpack::fastmks::FastMKSRules< KernelType, TreeType >::Score	(	const size_t	queryIndex,
		TreeType &	referenceNode
	)

Get the score for recursion order.

A low score indicates priority for recursion, while DBL_MAX indicates that the node should not be recursed into at all (it should be pruned).

Parameters

queryIndex	Index of query point.
referenceNode	Candidate to be recursed into.

Score() [2/2]

template<typename KernelType , typename TreeType >

double mlpack::fastmks::FastMKSRules< KernelType, TreeType >::Score	(	TreeType &	queryNode,
		TreeType &	referenceNode
	)

Get the score for recursion order.

A low score indicates priority for recursion, while DBL_MAX indicates that the node should not be recursed into at all (it should be pruned).

Parameters

queryNode	Candidate query node to be recursed into.
referenceNode	Candidate reference node to be recursed into.

---

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

• src/mlpack/methods/fastmks/fastmks_rules.hpp
• src/mlpack/methods/fastmks/fastmks_rules_impl.hpp