Learning from large, complex and structured data: advances in methods and applications

The community of the L’EMbeDS Department of Excellence (Economics, Management and Law in the era of Data Science) at the Sant’Anna School, and the broader community of FAIR’s Spoke 1 (Human-Centered AI) will come together for this two-day event held at the Sant’Anna School.

The workshop sessions will showcase diverse contributions by young researchers involved in L’EMbeDS, the SMaRT COnSTRUCT project funded by FAIR at Sant’Anna, and the PhD AI for Society – as well as research at the frontiers of statistical methodology and computer science. In addition, we will host keynotes by internationally renowned researchers in the fields of computer science, economics and geoinformatics.

All are welcome, and undergraduate and graduate students from all disciplines are particularly encouraged to participate. Details of the program are provided below.

Registration

Remote participation will be available (below the web meeting links).

The event is free, but please fill the registration form to let us know you are coming >>> link

Program

Tuesday June 3, 2025. Aula Magna, Sant’Anna School (sede centrale)

Link to join the web meeting (first day)

9:00-9:30: Welcome. Nicola Vitiello (Rector of the Sant’Anna School) and Francesca Chiaromonte (L’EMbeDS Scientific Coordinator, SMaRT COnSTRUCT co-PI).

9:30-10:30: Keynote: Symmetry reduction via linear invariant subspaces: exact and approximate methods. Mirco Tribastone, IMT Lucca – Chair: Andrea Vandin

10:30-11:00: Coffee break

11:00-13:00: Research from AI for Society (PhD-AI.it) – Chair: Riccardo Porcedda

• Riccardo Porcedda, Sant’Anna School. Generating Doppelganger Graphs: Preserving Privacy in Attributed, Weighted, and Directed Graphs.
• Vittoria Caponecchia (with Bernardo D'Agostino), Sant’Anna School. Process Mining for Legal Courts: Visualizing, Analyzing and Comparing Italian Divorce Proceedings.
• Lorenzo Emer, Sant’Anna School. Patent Analysis and Technology Forecasting with LLMs and Network Science.
• Luca Carmisciano, Sant’Anna School. Opportunities and Challenges of Functional Data in Medicine.
• Claudio Mazzi, Sant’Anna School. Machine Learning and Explainable AI for Type-2 Diabetes Management.

13:00-14:00: Lunch

14:00-16:00: Research from L’EMbeDS – Chair: Matteo Coronese

• Matteo Coronese, Sant’Anna School. Inferring Beliefs on Climate Risk from a Data-driven Agent-based Model of the Housing Market.
• Giulia Schneider, Sant’Anna School. Corporate Digital Responsibility: From Regulation of Responsible AI to Governance.
• Roberto Molinari, Auburn University. Fiducial Matching: Differentially Private Inference for Categorical Data.
• Valentina Lorenzoni, Sant’Anna School. Exploring the Truth about Age and Ageing.

16:00-16:30: Coffee break

16:30-18:30: The frontiers of Statistics – Chair: Lorenzo Testa

• Lorenzo Testa, Carnegie Mellon University. Semi-supervised Learning under Distribution Shift and Decaying Overlap.
• Ana Kenney, UC Irvine. Distilling Heterogeneous Treatment Effects: Stable Subgroup Estimation in Causal Inference.
• Luca Insolia, University of Geneva. Towards Optimal Adjustments for Multivariate Average Equivalence Testing.
• Jacopo Di Iorio, Emory University. Discovering Functional Motifs using funBIalign.
• Tobia Boschi, IBM Research Dublin. Functional Graph Convolutional Networks: A Unified Multi-task and Multi-modal Learning Framework.

18:30-19:30: Aperitif

Wednesday June 4, 2025. Aula Magna, Sant’Anna School (sede centrale)

Link to join the web meeting (second day)

9:00-10:00: Keynote: Learning for Data Structure Design. Paolo Ferragina, Sant’Anna School - Chair: Francesca Chiaromonte

10:00-10:30: Coffee break

10:30-12:30: Research from SMaRT COnSTRUCT – Chair: Simone Tonini

• Simone Tonini, Sant’Anna School. Fast, Robust and Accurate Anomaly Detection for Multivariate Time Series.
• Saverio Barabuffi, Sant’Anna School. Inter-regional Public Knowledge Spillovers and Patenting Performance: Evidence from US Metropolitan Statistical Areas through NLP Techniques.
• Irina Carnat, Sant’Anna School (online). Regulating Generative AI: What's the Purpose?
• Mario Martinoli, Sant’Anna School. Estimation of DSGE Models by non-Gaussian Vector Autoregressions.
• Michele Zema, Scuola Normale Superiore. Growing Through Links: evolution and impact of Production Networks in the long-run.

12:30-13:30: Lunch

13:30-14:30: Keynote: The Rise of Generative AI in Research and its Implications for Science. Philip Shapira, University of Manchester & Georgia Institute of Technology – Chair: Giulio Ferrigno

14:30-16:30: The frontiers of Computer Science – Chair: Andrea Vandin

• Emilio Incerto, IMT Lucca. Learning Stochastic Processes with Variable-length Markov Chains.
• Giuseppe Squillace, IMT Lucca. Efficient Network Embedding by Approximate Equitable Partitions.
• Francesco Tosoni, University of Pisa. Toward Greener Matrix Operations by Lossless Compressed Formats.
• Lorenzo Bellomo, University of Pisa. Dealing with Big Data: A Computation-Friendly Compressed Graph Database.
• Praveen Bushipaka, Sant’Anna School. Machine Unlearning in LLMs.

16:30-17:00:Coffee break

17:00-18:00: Keynote: The AI/ML Revolution and the Acceleration of Scientific Discovery in Earth Science. Guido Cervone, Penn State University (online) – Chair: Francesca Chiaromonte

18:00-18:30: Concluding remarks. Francesca Chiaromonte (L’EMbeDS Scientific Coordinator, SMaRT COnSTRUCT co-PI) and Andrea Vandin (SMaRT COnSTRUCT co-PI).