non-negative

On optimizing distributed non-negative Tucker decomposition

We discuss our experience in optimizing the non-negative Tucker decomposition for sparse datasets on a distributed system.