Non-Negative Tensor Factorization

Non-Negative Tensor Factorization is a multi-way extension of standard Non-Negative Matrix Factorization. See also: Non-negative tensor factorization with applications to statistics and computer vision.

Parameters

Parameter Name	Default Value	Description
numComponents	10	Number of components of the loading matrices.
gradientUpateType	GRADIENT_UPDATE_TYPE.NORMALIZED_UPDATE	Update method for the decomposed values. Can be one of GRADIENT_UPDATE_TYPE.{NORMALIZED_UPDATE, STEP_UPDATE_CUSTOM,ITERATION_UPDATE_CUSTOM}.
updater	none	Can be an implementation of org.nd4j.linalg.learning.config.IUpdater. Only used of gradientUpdateType is not GRADIENT_UPDATE_TYPE.NORMALIZED_UPDATE

Example Code

// Get data
Tensor X = Tensor.create(xdata);

// Setup model
int nComponents = ... // Choose a number of components F for the loading matrices
NTF ntf = new NTF();
ntf.setNumComponents(nComponents);

// Build  model
ntf.build(X);
Tensor[] decomposition = ntf.getDecomposition(); // Get #numModes matrices that represent the tensor decomposition