In this work we present and briefly review the Deep Unfolding, a recent machine learning paradigm that found natural application in scientific computing thanks to its promising strength in generating highly interpretable deep neural networks apt to be employed even on limited-resourced embedded systems. We describe this technique within a bilevel optimization framework and provide several exemplar applications, mainly focusing on Nonnegative Matrix Factorization.
Deep Unfolding for Scientific Computing on Embedded Systems
Chinellato E.;Marcuzzi F.
2026
Abstract
In this work we present and briefly review the Deep Unfolding, a recent machine learning paradigm that found natural application in scientific computing thanks to its promising strength in generating highly interpretable deep neural networks apt to be employed even on limited-resourced embedded systems. We describe this technique within a bilevel optimization framework and provide several exemplar applications, mainly focusing on Nonnegative Matrix Factorization.
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
