21/02 – seminario “Graph Signal Processing: Theory and Applications”

Il giorno 21 febbraio 2019 alle ore 10.00 presso l’Università degli Studi di Roma “Tor Vergata” Macroarea di Ingegneria – Via del Politecnico, 1 – Aula B15 – Edificio Didattico

Il Prof. Sergio Barbarossa dell’ Università di Roma La Sapienza terrà un seminario organizzato dal Prof. Gaspare Galati per la Ph.D. School in Electronic Engineering – DIE – Tor Vergata University su:

Graph Signal Processing: Theory and Applications

Abstract: Graph-based representations play a significant role in machine learning as a formal way to capture (pairwise) relations among data or time series. In this talk, we review the fundamentals of graphical methods applied to unsupervised and semisupervised learning and then we present some aspects of current research in the field, including the case where the knowledge about the graph is imperfect or when it is necessary to go beyond pairwise representations.

Sergio Barbarossa is a full professor in the Information Engineering, Electronics and Telecommunications Department of Sapienza University of Rome, Italy. He is an IEEE Fellow and EURASIP Fellow. For more than 25 years, he has been working in signal processing, radar remote sensing, sensor and telecommunication networks. Since 2000 he has been involved in European projects as a Technical Manager or Principal Investigator. He is now the technical manager of a H2020 Europe‐Japan project on 5G. His main current research interests are on mobile edge computing, 5G networks, distributed optimization, machine learning and signal processing on graphs.