Large scale optimization methods and applications in tensor optimizationLarge scale optimization methods and applications in tensor optimizationΤεχνικές βελτιστοποίησης μεγάλης κλίμακας με εφαρμογές σε tensors Μεταπτυχιακή Διατριβή Master Thesis 2017-12-142017enWe consider the problems of nonnegative tensor factorization and completion. Our aim is to derive efficient algorithms that are also suitable for parallel implementation. We adopt the alternating optimization framework and solve each matrix nonnegative least-squares problem via a Nesterov-type algorithm for convex and strongly convex problems. We describe parallel implementations of the algorithms and measure the attained speedup in a multi-core computing environment. It turns out that the derived algorithms are competitive candidates for the solution of very large-scale nonnegative tensor factorization and completion.http://creativecommons.org/licenses/by/4.0/Πολυτεχνείο Κρήτης::Σχολή Ηλεκτρολόγων Μηχανικών και Μηχανικών ΥπολογιστώνLourakis_Georgios_MSc_2017.pdfChania [Greece]Library of TUC2017-12-14application/pdf322.1 kBfree Lourakis Georgios Λουρακης Γεωργιος Liavas Athanasios Λιαβας Αθανασιος Digalakis Vasilis Διγαλακης Βασιλης Karystinos Georgios Καρυστινος Γεωργιος Πολυτεχνείο Κρήτης Technical University of Crete Parallel algorithms Optimal first-order optimization algorithms Nonnegative tensor factorization Tensors