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

Class URP

java.lang.Object

es.upm.etsisi.cf4j.recommender.Recommender

es.upm.etsisi.cf4j.recommender.matrixFactorization.URP

public class URP
extends Recommender

Implements Marlin, B. M. (2004). Modeling user rating profiles for collaborative filtering. In Advances in neural information processing systems (pp. 627-634).

Field Summary

Fields

Modifier and Type	Field	Description
protected double[]	alpha	Alpha parameter
protected double[][][]	beta	Beta parameter
protected static double	DEFAULT_H
protected static double	EPSILON
protected double[][]	gamma	Gamma parameter
protected double	H	Heuristic factor to control number of iterations during E-Step
protected int	numFactors	Number of latent factors
protected int	numIters	Number of iterations
protected Map<Integer,double[][]>	phi	Phi parameter
protected double[]	ratings	Plausible ratings (must be sorted in ascending order)

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

datamodel

Constructor Summary

Constructors

Constructor	Description
URP(DataModel datamodel, int numFactors, double[] ratings, int numIters)	Model constructor
URP(DataModel datamodel, int numFactors, double[] ratings, int numIters, double H)	Model constructor
URP(DataModel datamodel, int numFactors, double[] ratings, int numIters, double H, long seed)	Model constructor
URP(DataModel datamodel, int numFactors, double[] ratings, int numIters, long seed)	Model constructor
URP(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	getH()	Get the H value
int	getNumFactors()	Get the number of factors of the model
int	getNumIters()	Get the number of iterations
double[]	getPredictionProbabilityDistribution(int userIndex, int itemIndex)	Returns the probability distribution of a prediction.
double[]	getRatings()	Get the plausible ratings
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_H

protected static final double DEFAULT_H

See Also:

Constant Field Values

EPSILON

protected static final double EPSILON

See Also:

Constant Field Values

numIters

protected final int numIters

Number of iterations

numFactors

protected final int numFactors

Number of latent factors

H

protected final double H

Heuristic factor to control number of iterations during E-Step

ratings

protected final double[] ratings

Plausible ratings (must be sorted in ascending order)

gamma

protected final double[][] gamma

Gamma parameter

beta

protected final double[][][] beta

Beta parameter

alpha

protected final double[] alpha

Alpha parameter

phi

protected final Map<Integer,double[][]> phi

Phi parameter

Constructor Detail

URP

public URP(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.
• ratings: double array with the plausible ratings of the DataModel.
• numIters:: int value with the number of iterations.
• H (optional): double value that represents the heuristic factor to control the number of iterations during E-Step. The number of iterations is defined by H * number_of_user_ratings.
• seed (optional): random seed for random numbers generation. If missing, random value is used.

Parameters:

datamodel - DataModel instance

params - Model's hyper-parameters values

URP

public URP(DataModel datamodel,
           int numFactors,
           double[] ratings,
           int numIters)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of latent factors

ratings - Plausible ratings (must be sorted in ascending order)

numIters - Number of iterations

URP

public URP(DataModel datamodel,
           int numFactors,
           double[] ratings,
           int numIters,
           long seed)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of latent factors

ratings - Plausible ratings (must be sorted in ascending order)

numIters - Number of iterations

seed - Seed for random numbers generation

URP

public URP(DataModel datamodel,
           int numFactors,
           double[] ratings,
           int numIters,
           double H)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of latent factors

ratings - Plausible ratings (must be sorted in ascending order)

numIters - Number of iterations

H - Heuristic factor to control number of iterations during E-Step. The number of iterations is defined by H * number_of_user_ratings

URP

public URP(DataModel datamodel,
           int numFactors,
           double[] ratings,
           int numIters,
           double H,
           long seed)

Model constructor

Parameters:

datamodel - DataModel instance

numFactors - Number of latent factors

ratings - Plausible ratings (must be sorted in ascending order)

numIters - Number of iterations

H - Heuristic factor to control number of iterations during E-Step. The number of iterations is defined by H * number_of_user_ratings

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

getRatings

public double[] getRatings()

Get the plausible ratings

Returns:

Plausible ratings

getH

public double getH()

Get the H value

Returns:

H value

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

getPredictionProbabilityDistribution

public double[] getPredictionProbabilityDistribution(int userIndex,
                                                     int itemIndex)

Returns the probability distribution of a prediction. Each position of the returned array corresponds with the probability that the user (defined by userIndex) rates the item (defined by itemIndex) with the rating value that is in the same position in the array returned by getRatings()

Parameters:

userIndex - User

itemIndex - Item

Returns:

Probability distribution of the prediction

toString

public String toString()

Overrides:

toString in class Object