Package es.upm.etsisi.cf4j.recommender.matrixFactorization

Class BiasedMF

java.lang.Object

es.upm.etsisi.cf4j.recommender.Recommender

es.upm.etsisi.cf4j.recommender.matrixFactorization.BiasedMF

public class BiasedMF
extends Recommender

Implements Koren, Y., Bell, R., & Volinsky, C. (2009). Matrix factorization techniques for recommender systems. Computer, (8), 30-37.

Field Summary

Fields

Modifier and Type	Field	Description
protected double[]	bi	Item bias
protected double[]	bu	User bias
protected static double	DEFAULT_GAMMA
protected static double	DEFAULT_LAMBDA
protected double	gamma	Learning rate
protected double	lambda	Regularization parameter
protected int	numFactors	Number of latent factors
protected int	numIters	Number of iterations
protected double[][]	p	User factors
protected double[][]	q	Item factors

Fields inherited from class es.upm.etsisi.cf4j.recommender.Recommender

datamodel

Constructor Summary

Constructors

Constructor	Description
BiasedMF(DataModel datamodel, int numFactors, int numIters)	Model constructor
BiasedMF(DataModel datamodel, int numFactors, int numIters, double lambda)	Model constructor
BiasedMF(DataModel datamodel, int numFactors, int numIters, double lambda, double gamma, long seed)	Model constructor
BiasedMF(DataModel datamodel, int numFactors, int numIters, double lambda, long seed)	Model constructor
BiasedMF(DataModel datamodel, int numFactors, int numIters, long seed)	Model constructor
BiasedMF(DataModel datamodel, Map<String,Object> params)	Model constructor from a Map containing the model's hyper-parameters values.

Method Summary

Modifier and Type	Method	Description
void	fit()	Estimates model parameters given the hyper-parameters
double	getGamma()	Get the learning rate parameter of the model
double	getItemBias(int itemIndex)	Get the bias of a item (bi)
double[]	getItemFactors(int itemIndex)	Get the latent factors vector of an item (qi)
double	getLambda()	Get the regularization parameter of the model
int	getNumFactors()	Get the number of factors of the model
int	getNumIters()	Get the number of iterations
double	getUserBias(int userIndex)	Get the bias of a user (bu)
double[]	getUserFactors(int userIndex)	Get the latent factors vector of a user (pu)
double	predict(int userIndex, int itemIndex)	Computes a rating prediction
String	toString()

Methods inherited from class es.upm.etsisi.cf4j.recommender.Recommender

getDataModel, predict

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

DEFAULT_GAMMA

protected static final double DEFAULT_GAMMA

See Also:

Constant Field Values

DEFAULT_LAMBDA

protected static final double DEFAULT_LAMBDA

See Also:

Constant Field Values

p

protected final double[][] p

User factors

q

protected final double[][] q

Item factors

bu

protected final double[] bu

User bias

bi

protected final double[] bi

Item bias

gamma

protected final double gamma

Learning rate

lambda

protected final double lambda

Regularization parameter

numFactors

protected final int numFactors

Number of latent factors

numIters

protected final int numIters

Number of iterations

Constructor Detail

BiasedMF

public BiasedMF(DataModel datamodel,
                Map<String,Object> params)

Model constructor from a Map containing the model's hyper-parameters values. Map object must contains the following keys:

• numFactors: int value with the number of latent factors.
• numIters:: int value with the number of iterations.
• gamma (optional): double value with the learning rate hyper-parameter. If missing, it is set to 0.01.
• lambda (optional): double value with the regularization hyper-parameter. If missing, it is set to 0.05.
• seed (optional): random seed for random numbers generation. If missing, random value is used.

Parameters:

datamodel - DataModel instance

params - Model's hyper-parameters values

BiasedMF

public BiasedMF(DataModel datamodel,
                int numFactors,
                int numIters)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of factors

numIters - Number of iterations

BiasedMF

public BiasedMF(DataModel datamodel,
                int numFactors,
                int numIters,
                long seed)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of factors

numIters - Number of iterations

seed - Seed for random numbers generation

BiasedMF

public BiasedMF(DataModel datamodel,
                int numFactors,
                int numIters,
                double lambda)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of factors

numIters - Number of iterations

lambda - Regularization parameter

BiasedMF

public BiasedMF(DataModel datamodel,
                int numFactors,
                int numIters,
                double lambda,
                long seed)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of factors

numIters - Number of iterations

lambda - Regularization parameter

seed - Seed for random numbers generation

BiasedMF

public BiasedMF(DataModel datamodel,
                int numFactors,
                int numIters,
                double lambda,
                double gamma,
                long seed)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of factors

numIters - Number of iterations

lambda - Regularization parameter

gamma - Learning rate parameter

seed - Seed for random numbers generation

Method Detail

getNumFactors

public int getNumFactors()

Get the number of factors of the model

Returns:

Number of factors

getNumIters

public int getNumIters()

Get the number of iterations

Returns:

Number of iterations

getLambda

public double getLambda()

Get the regularization parameter of the model

Returns:

Lambda

getGamma

public double getGamma()

Get the learning rate parameter of the model

Returns:

Gamma

getUserFactors

public double[] getUserFactors(int userIndex)

Get the latent factors vector of a user (pu)

Parameters:

userIndex - User index

Returns:

Latent factors vector

getUserBias

public double getUserBias(int userIndex)

Get the bias of a user (bu)

Parameters:

userIndex - User index

Returns:

Bias

getItemFactors

public double[] getItemFactors(int itemIndex)

Get the latent factors vector of an item (qi)

Parameters:

itemIndex - Item index

Returns:

Latent factors vector

getItemBias

public double getItemBias(int itemIndex)

Get the bias of a item (bi)

Parameters:

itemIndex - Item index

Returns:

Bias

fit

public void fit()

Description copied from class: Recommender

Estimates model parameters given the hyper-parameters

Specified by:

fit in class Recommender

predict

public double predict(int userIndex,
                      int itemIndex)

Description copied from class: Recommender

Computes a rating prediction

Specified by:

predict in class Recommender

Parameters:

userIndex - Index of the user in the array of Users of the DataModel instance

itemIndex - Index of the item in the array of Items of the DataModel instance

Returns:

Prediction

toString

public String toString()

Overrides:

toString in class Object