edu.ucla.sspace.matrix

Interface MatrixFactorization

All Known Implementing Classes:

AbstractSvd, NonNegativeMatrixFactorizationMultiplicative, NonNegativeMatrixFactorizationOPL, SingularValueDecompositionLibC, SingularValueDecompositionLibJ, SingularValueDecompositionMatlab, SingularValueDecompositionOctave

public interface MatrixFactorization

An interface for any algorithm that decomposes a Matrix into two sub Matrixs that are linked by some form of latent variable. Examples of these decomposition algorithms are Singular Value Decomposition, Principal Component Analysis, and Non-Negative Matrix Factorization.
Given a MatrixFactorization algorithm, call factorize with a SparseMatrix representing the data points and the desired number of dimensions. The algorithm will then split the matrix into two smaller Matrixs: a data point by latent variable Matrix and a latent variable by feature Matrix. The interpretations of these matrices differer for each algorithm, but in all cases, these can be used as reduced representations of either the data points or the features. with a minimal loss of information.

Author:

Keith Stevens

Method Summary

Methods

Modifier and Type	Method and Description
Matrix	classFeatures() Returns the latent class by feature Matrix.
Matrix	dataClasses() Returns the data point by latent class Matrix.
void	factorize(MatrixFile m, int numDimensions) Factorizes the MatrixFile m.
void	factorize(SparseMatrix m, int numDimensions) Factorizes the SparseMatrix m.
MatrixBuilder	getBuilder() Returns a MatrixBuilder that is optimized for MatrixFiles that will be processed by this MatrixFactorization algorithm.

Method Detail

getBuilder

MatrixBuilder getBuilder()

Returns a MatrixBuilder that is optimized for MatrixFiles that will be processed by this MatrixFactorization algorithm.

factorize

void factorize(SparseMatrix m,
             int numDimensions)

Factorizes the SparseMatrix m. If m is of size MxN, this will generate two matrices: one of size MxnumDimensions and one of size numDimensionsxN. This method must be called before calling dataClasses and classFeatures.

factorize

void factorize(MatrixFile m,
             int numDimensions)

Factorizes the MatrixFile m. If m is of size MxN, this will generate two matrices: one of size MxnumDimensions and one of size numDimensionsxN. This method must be called before calling dataClasses and classFeatures.

dataClasses

Matrix dataClasses()

Returns the data point by latent class Matrix. This matrix represents the degree by which each data point can be explained by the discovered latent classes. Actual interpretations of the latent classes depends on the actual algorithm used. This Matrix can be used as a reduced representation of the data points themselves.

classFeatures

Matrix classFeatures()

Returns the latent class by feature Matrix. This matrix represents the degree by which each feature affects each latent class. Actual interpretations of this interaction depends on the actual algorithm used. This Matrix can be used as a reduced representation of the features.