Program
* You can book your place for the social dinner at the registration desk during the registration.
Keynote Speakers
Franco Scarselli
Theory of graph neural networks: some results and perspectives
By using graphs, we can naturally represent relationships together with patterns, namely domains where the information is composed by pieces of related data. From this point of view, machine learning methods for graphs are just an extension of standard models to more complex data. However, the presence of relationships and the fact that data complexity can be further increased suggests many novel application and research opportunities. Having the above idea in mind, in this talk, I will focus on theory of graph neural networks, recalling existing results and introducing some of those obtained in our lab. Moreover, I will discuss the perspectives and some of the open problems.
Alessandro Sperduti
Efficient Graph Neural Networks
Efficient Graph Neural Networks
We discuss different strategies to reduce the training burden of Graph Neural Networks with no or just marginal loss in performance. For node tasks we present a hierarchy of models based on simple graph convolution operators of increasing complexity that rely on linear transformations or controlled nonlinearities, and that can be implemented in single-layer graph convolutional networks. We also introduce a novel convolutional operator named Compact Multi-head EGC (CM-EGC) that beside exploiting a very simple graph convolution definition, also significantly reduces the number of learnable parameters compared to existing convolutions.Another applicable strategy consists in exploiting a reservoir architecture. Multiresolution Reservoir Graph Neural Networks (MRGNNs), inspired by graph spectral filtering, are an example of such approach. They are based on an explicit k-hop unsupervised graph representation amenable for further nonlinear processing. On this line, we also report on results obtained by untrained Graph Neural Networks for fast and accurate graph classification. Finally, we present a Backpropagation-Free training algorithm that allows to achieve competitive results for node classification tasks, while considerably reducing the training burden.
Public Talk
Giulia Cencetti
What is Network Science?
What is Network Science?
Everybody talks about networks, networking, connections, and complexity, but are we sure we know what we are talking about? In this seminar we will try to understand together what networks are, and why we cannot talk about networks without talking about complex systems. We will discover the interdisciplinary power of this subject. Then I will give a taste of how to cope with networks: How to find structures, identify common patterns, detect similarities and differences. We will talk about centrality, six degrees of separation (and why it’s a small world), communities and higher-order networks. We will see how all this helps in explaining blackouts, managing epidemics, spreading climate change knowledge, and also understanding history. To conclude, we will focus on the role of women in network science.
The public discussion is scheduled to take place at "La Bookique" located at Via Torre d'Augusto, 29, 38122 Trento TN, on Wednesday, the 29th, at 9:00 pm.
This event is open to everyone.
Poster
Petya Vasileva (University of Michigan)
Graph Neural Networks for Problem Detection in Scientific Network Infrastructures
Francesco Paolo Nerini (Centai Institute)
Value is in the Eye of the Beholder: A Framework for an Equitable Graph Data EvaluationMarco Vincenzo De Luca (University of Trento)
xAI-based Regularizers for Graph Neural NetworksVeronica Lachi (University of Siena)
The Expressive Power of Pooling in Graph Neural NetworksSimone Piaggesi (University of Pisa)
DINE: Dimensional Interpretability of Node EmbeddingsMarco Bronzini (University of Trento)
Glitter or Gold? Deriving Structured Insights from Sustainability Reports via Large Language Models
Donato Crisostomi (Sapienza, University of Rome)
Metric Based Few-Shot Graph ClassificationManuel Dileo (University of Milan)
DURENDAL: Graph deep learning framework for temporal heterogeneous networksAndrea Giuseppe Di Francesco (Sapienza University of Rome)
Link Prediction with Physics-Inspired Graph Neural Networks
Francesco Ferrini (University of Trento)
Meta-Path Learning for Multi-relational Graph Neural Networks
Giuseppe Alessio D'inverno (University of Siena)
VC dimension of Graph Neural Networks with Pfaffian activation functionsEffrosyni Papanastasiou (Sorbonne University & University of Siena)
Self-Supervised Directed Graph Structure Learning given Temporal Node InteractionsQuintino Francesco Lotito (University of Trento)
Hypergraphx: a library for higher-order network analysisCaterina Graziani (University of Siena)
PAIN: Expressive GNNs with Path AggregationJules Morand (University of Trento)
The lock-down communities management: Optimal spatio-temporal clustering of the inter-provincial Italian network during Covid-19 crisisMaria Sofia Bucarelli (Sapienza University of Rome)
Hypergraph Neural Networks through the Lens of Message Passing: A Common Perspective to Homophily and Architecture Design
Marco Pacini (Fondazione Bruno Kessler)
A characterization theorem for equivariant networks with point-wise activations.
Talks
Pietro Barbiero - (Universita' della Svizzera Italiana)
Interpretable Graph Networks Formulate Universal Algebra ConjecturesManuel Dileo - (Università degli Studi di Milano)
Blockchain-data: a playground for temporal graph learningDonato Crisostomi - (Sapienza Università di Roma)
Unifying RepresentationsDanilo Numeroso - (Università di Pisa)
Targeting Combinatorial Optimisation Problems through Neural Algorithmic ReasoningMatteo Tiezzi - (Università di Siena)
Graph Neural Networks for Graph DrawingIndro Spinelli - (Sapienza Università di Roma)
Designing Explainable Graph Neural NetworkPietro Barbiero - (Universita' della Svizzera Italiana)
Relational Concept Based Models