Package org.dulab.javanmf.algorithms

Provides classes for performing non-negative matrix factorization, non-negative matrix regression, and non-negative singular value decomposition.

• Non-negative matrix factorization (NMF), performed by alternating updates of matrices W and H to minimize the distance between X and WH.
• Non-negative one-matrix optimization, performed by updating matrix H to minimize the distance between X and WH.
• Non-negative singular value decomposition (NNDSVD), used to initialize matrix W and H. Based on C. Boutsidis and E. Gallopoulos, SVD based initialization: A head start for nonnegative matrix factorization].

Author:

Du-Lab Team dulab.binf@gmail.com

See Also:

MatrixFactorization, MatrixRegression, SingularValueDecomposition

Interface Summary

Interface	Description
Constraint

Class Summary

Class	Description
AlternatingLeastSquaresMatrixFactorization	This class performs non-negative matrix factorization using the alternating non-negative least squares method.
DefaultConstraint
MatrixFactorization	This class performs non-negative matrix factorization: for given matrix X, find matrices W and H that minimize the objective function
MatrixRegression	This class performs non-negative matrix regression: for given matrices X and W, find matrix H that minimizes the objective function
MatrixUtils
NonNegativeLeastSquares	This class solves the non-negative least squares problem using the active set method.
SingularValueDecomposition	This class performs non-negative singular value decomposition: first, the singular value decomposition is performed; then, the non-negative matrices W and H are formed.