Dr. Lodovico Giaretta (RISE) was the first expert to visit UNIZG-FER from 8th to 10th November 2023 as part of the AIoTwin activities. Lodovico presented a series of three lectures with a special focus on Graph Neural Networks (GNNs) and decentralized systems and shed light on these cutting-edge technologies.
On November 8th, he gave a public talk entitled "Causal Temporal GNNs as Decentralized Memory Networks," , co-organized by the IEEE Croatian Section, which was open to the general public to provide a broad audience with a glimpse into the evolving landscape of GNNs and decentralized systems.
In contrast, the following two presentations were tailored for members of the AIoTwin project, offering a more specific exploration of these cutting-edge technologies.
- Decentralised Data Mining using Gossip Protocols: This lecture introduced the broad family of Gossip Protocols and presented a detailed analysis of a few key use cases and specific techniques.
- Graph Representation Learning with Graph Neural Networks: This lecture introduced the broad topic of Graph Representation Learning (GRL), with a particular focus on Graph Neural Network (GNN) techniques.
During the lectures Lodovico explored the foundational principles of GNNs, their diverse applications, and the inherent challenges associated with decentralized systems. Attendees gained a better understanding of the dynamic nature of graphs, the implications of temporal changes on algorithms, and the intricacies of decentralized inference in GNNs.