MARCO BENEDETTI

Dottore di ricerca

ciclo: XXXV

Titolo della tesi: Models for Associative Memory in Artificial and Biological Neural Networks

Since it's onset, research in the field of Artificial Neural Networks has pursued two different, but far from alternative objectives. On the one hand, there is a strong interest in modelling the functioning of a biological brain. On the other hand, there is a very practical interest in harvesting the secrets of the brain's extraordinary efficiency and precision in data elaboration, which is in the general case far from matched by state of the art Artificial Intelligence. In this thesis we discuss some interesting examples of this duality, in the context of Associative Memory, between algorithms and biologically inspired learning rules.

Produzione scientifica

11573/1706048 - 2024 - Eigenvector dreaming

Benedetti, Marco; Carillo, Louis; Marinari, Enzo; Mézard, Marc - 01a Articolo in rivista

rivista: JOURNAL OF STATISTICAL MECHANICS: THEORY AND EXPERIMENT (Bristol : IOP Publishing) pp. 1-16 - issn: 1742-5468 - wos: WOS:001143753800001 (0) - scopus: 2-s2.0-85182779151 (1)

11573/1680161 - 2022 - Supervised perceptron learning vs unsupervised Hebbian unlearning: Approaching optimal memory retrieval in Hopfield-like networks

Benedetti, Marco; Ventura, Enrico; Marinari, Enzo; Ruocco, Giancarlo; Zamponi, Francesco - 01a Articolo in rivista

rivista: THE JOURNAL OF CHEMICAL PHYSICS (American Institute of Physics:2 Huntington Quadrangle, Suite 1NO1:Melville, NY 11747:(800)344-6902, (631)576-2287, EMAIL: subs@aip.org, INTERNET: http://www.aip.org, Fax: (516)349-9704) pp. 104107- - issn: 0021-9606 - wos: WOS:000892834700012 (8) - scopus: 2-s2.0-85126701753 (9)

11573/1577265 - 2021 - Recognition capabilities of a Hopfield model with auxiliary hidden neurons

Benedetti, M.; Dotsenko, V.; Fischetti, G.; Marinari, E.; Oshanin, G. - 01a Articolo in rivista

rivista: PHYSICAL REVIEW. E (Ridge, NY: American Physical Society, [2016]-) pp. - - issn: 2470-0045 - wos: WOS:000661856700009 (1) - scopus: 2-s2.0-85108501063 (1)