item attribute to latent factor mapping, optimized for BPR loss More...

Inheritance diagram for BPRMF_ItemMapping_Optimal:

Public Member Functions
override void	AddFeedback (int user_id, int item_id)
	Add a positive feedback event and perform incremental training.
virtual bool	CanPredict (int user_id, int item_id)
	Check whether a useful prediction (i.e. not using a fallback/default answer) can be made for a given user-item combination.
Object	Clone ()
	create a shallow copy of the object
override float	ComputeObjective ()
	Compute the current optimization objective (usually loss plus regularization term) of the model.
override void	Iterate ()
	Perform one iteration of stochastic gradient ascent over the training data.
override void	IterateMapping ()
	Perform one iteration of the mapping learning process.
override void	LearnAttributeToFactorMapping ()
	Learn the mapping.
override void	LoadModel (string file)
	Get the model parameters from a file.
override float	Predict (int user_id, int item_id)
	Predict rating or score for a given user-item combination.
override void	RemoveFeedback (int user_id, int item_id)
	Remove all feedback events by the given user-item combination.
override void	RemoveItem (int item_id)
	Remove all feedback by one item.
override void	RemoveUser (int user_id)
	Remove all feedback by one user.
override void	SaveModel (string file)
	Save the model parameters to a file.
override string	ToString ()
	Return a string representation of the recommender.
override void	Train ()
	Learn the model parameters of the recommender from the training data.
Protected Member Functions
override float[]	__MapToLatentFactorSpace (int item_id)
	map to latent factor space (actual function)
override void	AddItem (int item_id)
override void	AddUser (int user_id)
void	CheckSampling ()
double[]	ComputeMappingFit ()
	Compute the fit of the mapping.
override void	InitModel ()
override float[]	MapToLatentFactorSpace (int item_id)
	map to latent factor space (method to be called)
virtual void	RetrainItem (int item_id)
	Retrain the latent factors of a given item.
virtual void	RetrainUser (int user_id)
	Retrain the latent factors of a given user.
int	SampleItem ()
	Samples an item for the mapping training. Only items that are associated with at least one user are taken into account.
virtual void	SampleItemPair (int u, out int i, out int j)
	Sample a pair of items, given a user.
virtual bool	SampleOtherItem (int u, int i, out int j)
	Sample another item, given the first one and the user.
virtual void	SampleTriple (out int u, out int i, out int j)
	Sample a triple for BPR learning.
virtual int	SampleUser ()
	Sample a user that has viewed at least one and not all items.
virtual void	UpdateFactors (int u, int i, int j, bool update_u, bool update_i, bool update_j)
	Update latent factors according to the stochastic gradient descent update rule.
virtual void	UpdateMappingFactors (int u, int i, int j)
	update the mapping factors for a given user and an item pair
Protected Attributes
Func< int, float[]>	_MapToLatentFactorSpace
	map to latent factor space (field)
Matrix< float >	attribute_to_factor
	The matrix representing the attribute-to-factor mapping.
float[]	factor_bias
	array to store the bias for each mapping
bool	fast_sampling = false
	Fast, but memory-intensive sampling.
int	fast_sampling_memory_limit = 1024
	Fast sampling memory limit, in MiB.
SparseBooleanMatrix	item_attributes
	The matrix storing the item attributes.
float[]	item_bias
	Item bias terms.
Matrix< float >	item_factors
	Latent item factor matrix.
float	learn_rate = 0.05f
	Learning rate alpha.
double	learn_rate_mapping = 0.01
	The learn rate for training the mapping functions.
int	num_factors = 10
	Number of latent factors per user/item.
int	num_init_mapping = 5
	number of times the regression is computed (to avoid local minima)
int	num_iter_mapping = 10
	number of iterations of the mapping training procedure
System.Random	random
	Random number generator.
float	reg_i = 0.0025f
	Regularization parameter for positive item factors.
float	reg_j = 0.00025f
	Regularization parameter for negative item factors.
double	reg_mapping = 0.1
	regularization constant for the mapping
float	reg_u = 0.0025f
	Regularization parameter for user factors.
bool	update_j = true
	If set (default), update factors for negative sampled items during learning.
Matrix< float >	user_factors
	Latent user factor matrix.
IList< IList< int > >	user_neg_items
	support data structure for fast sampling
IList< IList< int > >	user_pos_items
	support data structure for fast sampling
Properties
float	BiasReg `[get, set]`
	Regularization parameter for the bias term.
int	FastSamplingMemoryLimit `[get, set]`
	Fast sampling memory limit, in MiB.
virtual IPosOnlyFeedback	Feedback `[get, set]`
	the feedback data to be used for training
double	InitMean `[get, set]`
	Mean of the normal distribution used to initialize the latent factors.
double	InitStdDev `[get, set]`
	Standard deviation of the normal distribution used to initialize the latent factors.
SparseBooleanMatrix	ItemAttributes `[get, set]`
float	LearnRate `[get, set]`
	Learning rate alpha.
double	LearnRateMapping `[get, set]`
	The learn rate for training the mapping functions.
int	MaxItemID `[get, set]`
	Maximum item ID.
int	MaxUserID `[get, set]`
	Maximum user ID.
uint	NumFactors `[get, set]`
	Number of latent factors per user/item.
int	NumInitMapping `[get, set]`
	number of times the regression is computed (to avoid local minima)
int	NumItemAttributes `[get, set]`
uint	NumIter `[get, set]`
	Number of iterations over the training data.
int	NumIterMapping `[get, set]`
	number of iterations of the mapping training procedure
float	RegI `[get, set]`
	Regularization parameter for positive item factors.
float	RegJ `[get, set]`
	Regularization parameter for negative item factors.
double	RegMapping `[get, set]`
	regularization constant for the mapping
float	RegU `[get, set]`
	Regularization parameter for user factors.
bool	UniformUserSampling `[get, set]`
	Sample uniformly from users.
bool	UpdateJ `[get, set]`
	If set (default), update factors for negative sampled items during learning.
bool	WithReplacement `[get, set]`
	Sample positive observations with (true) or without (false) replacement.

Detailed Description

item attribute to latent factor mapping, optimized for BPR loss

Literature:

Zeno Gantner, Lucas Drumond, Christoph Freudenthaler, Steffen Rendle, Lars Schmidt-Thieme: Learning Attribute-to-Feature Mappings for Cold-Start Recommendations. ICDM 2011. http://www.ismll.uni-hildesheim.de/pub/pdfs/Gantner_et_al2010Mapping.pdf

This recommender does NOT support incremental updates.

Member Function Documentation

override float [] __MapToLatentFactorSpace ( int item_id ) [inline, protected, virtual]

map to latent factor space (actual function)

Reimplemented from BPRMF_ItemMapping.

override void AddFeedback	(	int	user_id,
		int	item_id
	)		`[inline, virtual, inherited]`

Add a positive feedback event and perform incremental training.

Parameters:

user_id	the user ID
item_id	the item ID

Reimplemented from IncrementalItemRecommender.

virtual bool CanPredict	(	int	user_id,
		int	item_id
	)		`[inline, virtual, inherited]`

Check whether a useful prediction (i.e. not using a fallback/default answer) can be made for a given user-item combination.

It is up to the recommender implementor to decide when a prediction is useful, and to document it accordingly.

Parameters:

user_id	the user ID
item_id	the item ID

Returns:: true if a useful prediction can be made, false otherwise

Implements IRecommender.

Object Clone ( ) [inline, inherited]

create a shallow copy of the object

double [] ComputeMappingFit ( ) [inline, protected, inherited]

Compute the fit of the mapping.

Returns:: an array of doubles containing the RMSE on the training data for each latent factor

override float ComputeObjective ( ) [inline, virtual, inherited]

Compute the current optimization objective (usually loss plus regularization term) of the model.

Returns:: the current objective; -1 if not implemented

Implements MF.

Reimplemented in SoftMarginRankingMF.

override void Iterate ( ) [inline, virtual, inherited]

Perform one iteration of stochastic gradient ascent over the training data.

One iteration is samples number of positive entries in the training matrix times

Implements MF.

Reimplemented in BPRMF_Mapping, and MultiCoreBPRMF.

override void IterateMapping ( ) [inline, virtual]

Perform one iteration of the mapping learning process.

Reimplemented from BPRMF_ItemMapping.

override void LearnAttributeToFactorMapping ( ) [inline, virtual]

Learn the mapping.

Reimplemented from BPRMF_ItemMapping.

override void LoadModel ( string filename ) [inline, virtual, inherited]

Get the model parameters from a file.

Parameters:

filename the name of the file to read from

Reimplemented from MF.

override float [] MapToLatentFactorSpace ( int item_id ) [inline, protected, virtual]

map to latent factor space (method to be called)

Reimplemented from BPRMF_ItemMapping.

override float Predict	(	int	user_id,
		int	item_id
	)		`[inline, inherited]`

Predict rating or score for a given user-item combination.

Parameters:

user_id	the user ID
item_id	the item ID

Returns:: the predicted score/rating for the given user-item combination

Implements IRecommender.

override void RemoveFeedback	(	int	user_id,
		int	item_id
	)		`[inline, virtual, inherited]`

Remove all feedback events by the given user-item combination.

Parameters:

user_id	the user ID
item_id	the item ID

Reimplemented from IncrementalItemRecommender.

override void RemoveItem ( int item_id ) [inline, virtual, inherited]

Remove all feedback by one item.

Parameters:

item_id the item ID

Reimplemented from IncrementalItemRecommender.

override void RemoveUser ( int user_id ) [inline, virtual, inherited]

Remove all feedback by one user.

Parameters:

user_id the user ID

Reimplemented from IncrementalItemRecommender.

virtual void RetrainItem ( int item_id ) [inline, protected, virtual, inherited]

Retrain the latent factors of a given item.

Parameters:

item_id the item ID

virtual void RetrainUser ( int user_id ) [inline, protected, virtual, inherited]

Retrain the latent factors of a given user.

Parameters:

user_id the user ID

int SampleItem ( ) [inline, protected, inherited]

Samples an item for the mapping training. Only items that are associated with at least one user are taken into account.

Returns:: the item ID

virtual void SampleItemPair	(	int	u,
		out int	i,
		out int	j
	)		`[inline, protected, virtual, inherited]`

Sample a pair of items, given a user.

Parameters:

u	the user ID
i	the ID of the first item
j	the ID of the second item

virtual bool SampleOtherItem	(	int	u,
		int	i,
		out int	j
	)		`[inline, protected, virtual, inherited]`

Sample another item, given the first one and the user.

Parameters:

u	the user ID
i	the ID of the given item
j	the ID of the other item

Returns:: true if the given item was already seen by user u

virtual void SampleTriple	(	out int	u,
		out int	i,
		out int	j
	)		`[inline, protected, virtual, inherited]`

Sample a triple for BPR learning.

Parameters:

u	the user ID
i	the ID of the first item
j	the ID of the second item

Reimplemented in WeightedBPRMF.

virtual int SampleUser ( ) [inline, protected, virtual, inherited]

Sample a user that has viewed at least one and not all items.

Returns:: the user ID

override void SaveModel ( string filename ) [inline, virtual, inherited]

Save the model parameters to a file.

Parameters:

filename the name of the file to write to

Reimplemented from MF.

override string ToString ( ) [inline]

Return a string representation of the recommender.

The ToString() method of recommenders should list the class name and all hyperparameters, separated by space characters.

Reimplemented from BPRMF_ItemMapping.

virtual void UpdateFactors	(	int	u,
		int	i,
		int	j,
		bool	update_u,
		bool	update_i,
		bool	update_j
	)		`[inline, protected, virtual, inherited]`

Update latent factors according to the stochastic gradient descent update rule.

Parameters:

u	the user ID
i	the ID of the first item
j	the ID of the second item
update_u	if true, update the user latent factors
update_i	if true, update the latent factors of the first item
update_j	if true, update the latent factors of the second item

Reimplemented in SoftMarginRankingMF.

virtual void UpdateMappingFactors	(	int	u,
		int	i,
		int	j
	)		`[inline, protected, virtual]`

update the mapping factors for a given user and an item pair

Parameters:

u	the user ID
i	the first item ID
j	the second item ID

Member Data Documentation

Func<int, float[]> _MapToLatentFactorSpace [protected, inherited]

map to latent factor space (field)

Matrix<float> attribute_to_factor [protected, inherited]

The matrix representing the attribute-to-factor mapping.

includes bias

float [] factor_bias [protected, inherited]

array to store the bias for each mapping

bool fast_sampling = false [protected, inherited]

Fast, but memory-intensive sampling.

int fast_sampling_memory_limit = 1024 [protected, inherited]

Fast sampling memory limit, in MiB.

SparseBooleanMatrix item_attributes [protected, inherited]

The matrix storing the item attributes.

float [] item_bias [protected, inherited]

Item bias terms.

Matrix<float> item_factors [protected, inherited]

Latent item factor matrix.

float learn_rate = 0.05f [protected, inherited]

Learning rate alpha.

double learn_rate_mapping = 0.01 [protected, inherited]

The learn rate for training the mapping functions.

int num_factors = 10 [protected, inherited]

Number of latent factors per user/item.

int num_init_mapping = 5 [protected, inherited]

number of times the regression is computed (to avoid local minima)

may be ignored by the recommender

int num_iter_mapping = 10 [protected, inherited]

number of iterations of the mapping training procedure

System.Random random [protected, inherited]

Random number generator.

float reg_i = 0.0025f [protected, inherited]

Regularization parameter for positive item factors.

float reg_j = 0.00025f [protected, inherited]

Regularization parameter for negative item factors.

double reg_mapping = 0.1 [protected, inherited]

regularization constant for the mapping

float reg_u = 0.0025f [protected, inherited]

Regularization parameter for user factors.

bool update_j = true [protected, inherited]

If set (default), update factors for negative sampled items during learning.

Matrix<float> user_factors [protected, inherited]

Latent user factor matrix.

IList<IList<int> > user_neg_items [protected, inherited]

support data structure for fast sampling

IList<IList<int> > user_pos_items [protected, inherited]

support data structure for fast sampling

Property Documentation

float BiasReg [get, set, inherited]

Regularization parameter for the bias term.

int FastSamplingMemoryLimit [get, set, inherited]

Fast sampling memory limit, in MiB.

virtual IPosOnlyFeedback Feedback [get, set, inherited]

the feedback data to be used for training

double InitMean [get, set, inherited]

Mean of the normal distribution used to initialize the latent factors.

double InitStdDev [get, set, inherited]

Standard deviation of the normal distribution used to initialize the latent factors.

SparseBooleanMatrix ItemAttributes [get, set, inherited]

the binary item attributes

Implements IItemAttributeAwareRecommender.

float LearnRate [get, set, inherited]

Learning rate alpha.

double LearnRateMapping [get, set, inherited]

The learn rate for training the mapping functions.

int MaxItemID [get, set, inherited]

Maximum item ID.

int MaxUserID [get, set, inherited]

Maximum user ID.

uint NumFactors [get, set, inherited]

Number of latent factors per user/item.

int NumInitMapping [get, set, inherited]

number of times the regression is computed (to avoid local minima)

may be ignored by the recommender

int NumItemAttributes [get, set, inherited]

an integer stating the number of attributes

Implements IItemAttributeAwareRecommender.

uint NumIter [get, set, inherited]

Number of iterations over the training data.

Implements IIterativeModel.

int NumIterMapping [get, set, inherited]

number of iterations of the mapping training procedure

float RegI [get, set, inherited]

Regularization parameter for positive item factors.

float RegJ [get, set, inherited]

Regularization parameter for negative item factors.

double RegMapping [get, set, inherited]

regularization constant for the mapping

float RegU [get, set, inherited]

Regularization parameter for user factors.

bool UniformUserSampling [get, set, inherited]

Sample uniformly from users.

bool UpdateJ [get, set, inherited]

If set (default), update factors for negative sampled items during learning.

bool WithReplacement [get, set, inherited]

Sample positive observations with (true) or without (false) replacement.

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

BPRMF_ItemMapping_Optimal.cs

Public Member Functions

Protected Member Functions

Protected Attributes

Properties

Detailed Description

Member Function Documentation

Member Data Documentation

Property Documentation