“For now, what is important is not finding the answer, but looking for it.”
I am fascinated by emergent phenomena: how simple rules give rise to complex behavior, in vitro as well as in silico. Today, artificial intelligence is the best playground to study them, and it raises the question I find most intriguing: is intelligence rooted in the physical world and in interacting with it, or can it emerge from understanding the causal relationships between symbols alone?
At AIRLab, Politecnico di Milano, I explore this question and many others, with a particular focus on AI for medicine. I also love teaching and mentoring, and I take part in educational initiatives to promote AI literacy at all levels.
Research interests
Tabular Foundation Models
One pretrained model for any table: predictions on a new dataset in a single forward pass, with no task-specific training.
Survival Analysis
Predicting not only whether an event happens but when, from patient outcomes to financial risk.
Federated Learning
Training models across hospitals without exposing patient data.
Machine Learning for Healthcare
Models clinicians can rely on: private by design, interpretable, and robust to messy multimodal data.
News
- Sep 2026
New preprint: Tabular Foundation Models: A Systematic Survey.
- Jul 2026
New preprint with Niccolò Maria Rizzi: A Filtered Mixture-of-Generators for Fully Synthetic Survival Training.
- Jun 2026
Presented SurvKAN and Deep Variational Contrastive Learning at IJCNN 2026 in Maastricht. Both grew out of MSc theses I co-supervised.
- Feb 2026
New preprint: QuAIL: Quality-Aware Inertial Learning for Robust Training under Data Corruption.
Publications
- 2026 Conference
Deep Variational Contrastive Learning for Joint Risk Stratification and Time-to-Event Estimation
Erbil, P., Archetti, Alberto, Lomurno, E., Matteucci, M.
2026 International Joint Conference on Neural Networks (IJCNN)
- 2026 Conference
SurvKAN: A Fully Parametric Survival Model Based on Kolmogorov-Arnold Networks
Mastroleo, M., Archetti, Alberto, Mastroleo, F., Matteucci, M.
2026 International Joint Conference on Neural Networks (IJCNN)
- 2026 Preprint
QuAIL: Quality-Aware Inertial Learning for Robust Training under Data Corruption
Sabella, M., Archetti, Alberto, Pinoli, P., Matteucci, M., Cappiello, C.
arXiv ↗ 1 citations - 2026 Preprint
A Filtered Mixture-of-Generators for Fully Synthetic Survival Training
Rizzi, N. M., Lomurno, E., Archetti, Alberto, Matteucci, M.
- 2026 Preprint
Tabular Foundation Models: A Systematic Survey
Archetti, Alberto, Mastroleo, M., Sabella, M., Cappiello, C., Matteucci, M.
SSRN Preprint
- 2025 Conference
FPBoost: Fully Parametric Gradient Boosting for Survival Analysis
Archetti, Alberto, Lomurno, E., Piccinotti, D., Matteucci, M.
European Conference on Artificial Intelligence (ECAI)
DOI ↗ 9 citations - 2025 Journal
Archetti, Alberto, Giusti, G., Gorla, K. R., Caputo, S., Magarini, M., Matteucci, M., Mucchi, L., Pierobon, M.
IEEE Transactions on Molecular, Biological, and Multi-Scale Communications
DOI ↗ 4 citations - 2024 Journal
Bridging the gap: improve neural survival models with interpolation techniques
Archetti, Alberto, Stranieri, F., Matteucci, M.
Progress in Artificial Intelligence
DOI ↗ 6 citations - 2024 Conference
Latent neural cellular automata for resource-efficient image restoration
Menta, A., Archetti, Alberto, Matteucci, M.
Artificial Life Conference Proceedings 36
DOI ↗ 8 citations - 2023 Journal
Archetti, Alberto, Ieva, F., Matteucci, M.
Future Generation Computer Systems
DOI ↗ 30 citations - 2023 Workshop
Heterogeneous Datasets for Federated Survival Analysis Simulation
Archetti, Alberto, Lomurno, E., Lattari, F., Martin, A., Matteucci, M.
Companion of the 2023 ACM/SPEC International Conference on Performance Engineering
DOI ↗ 20 citations - 2023 Conference
Archetti, Alberto, Matteucci, M.
2023 International Joint Conference on Neural Networks (IJCNN)
DOI ↗ 37 citations - 2023 Workshop
Deep Survival Analysis for Healthcare: An Empirical Study on Post-Processing Techniques
Archetti, Alberto, Stranieri, F., Matteucci, M.
CEUR Workshop Proceedings
4 citations - 2023 Conference
Discriminative adversarial privacy: balancing accuracy and membership privacy in neural networks
Lomurno, E., Archetti, Alberto, Ausonio, F., Matteucci, M.
The 34th British Machine Vision Conference Proceedings (BMVC)
8 citations - 2023 Conference
SGDE: Secure Generative Data Exchange for Cross-Silo Federated Learning
Lomurno, E., Archetti, Alberto, Cazzella, L., Samele, S., Di Perna, L., Matteucci, M.
Proceedings of the 2022 5th International Conference on Artificial Intelligence and Pattern Recognition
DOI ↗ 34 citations - 2023 Workshop
Drug Inventory Control: Human Decisions versus Deep Reinforcement Learning
Stranieri, F., Archetti, Alberto, Robbiano, E., Kouki, C., Stella, F.
CEUR Workshop Proceedings
- 2022 Conference
Neural Weighted A*: Learning Graph Costs and Heuristics with Differentiable Anytime A*
Archetti, Alberto, Cannici, M., Matteucci, M.
Machine Learning, Optimization, and Data Science (LOD 2021)
DOI ↗ 18 citations
Experience & Education
-
Research Fellow
2025 – present AIRLab, Politecnico di Milano · Milan, ItalyResearch topic: Tabular foundation models for healthcare.
-
National Ph.D. in Artificial Intelligence for Industry 4.0
2021 – 2025 Politecnico di TorinoResearch topic: Federated Learning for Survival Analysis.
Collaborations: Research visitor at the Human Technopole, Milan, Italy.
Thesis: Federated Survival Analysis: Ensemble and Neural Methods for Distributed Time-to-Event Data
-
Research Internship
2021 Politecnico di Milano · Milan, ItalyActivities: Data analysis for anomaly detection; automated knowledge graph extraction.
-
Master of Science in Computer Science and Engineering
2018 – 2021 Politecnico di Milano · 110/110 cum LaudeThesis: Neural Weighted A*: Learning Graph Costs and Heuristics with Differentiable Anytime A*
-
Bachelor's Degree in Computer Science and Engineering
2015 – 2018 Politecnico di Milano · 110/110 cum Laude -
Diploma di Liceo Scientifico
2010 – 2015 Madonna della Neve · 100/100
Research projects
AI-SPRINT
Artificial Intelligence in Secure PRIvacy-preserving computing coNTinuum
AI scientific lead
Teaching & Mentoring
Courses 8 courses · 351 h
Computer Grafica per il Game Design
Professor
Artificial Intelligence for Business
Professor
Advanced Deep Learning
Teaching Assistant
Artificial Neural Networks and Deep Learning
Teaching Assistant · 3 editions
AI Bootcamp
Teaching Assistant · 3 editions
AI Product Management Bootcamp
Teaching Assistant
Coding Bootcamp
Teaching Assistant · 3 editions
Software Engineering
Lab Tutor · 3 editions
Thesis supervision 9 theses
Pinar Erbil
Deep Variational Contrastive Learning for Risk Stratification and Time-to-Event Estimation
Marina Mastroleo
SurvKAN: Fully Parametric Survival Modeling with Kolmogorov-Arnold Networks
Sofia Perini
Exploring multimodality in federated survival analysis
Emanuele Paesano
A multimodal framework for survival analysis integrating clinical, genomic, histopathological and textual data
Niccolò Maria Rizzi
A generative pipeline for high-quality synthetic survival datasets
Gabriele Giusti
Multi-Agent Reinforcement Learning for emergent molecular communication in diffusion-based environments
Andrea Menta
Image restoration via Latent Neural Cellular Automata
Simone Cimmino
A pipeline for company industrial sector classification from unstructured website content
Sara Sacco
Analisi spettrale per il rilevamento di anomalie: studio sulla sopravvivenza degli estensimetri a corda per Snam S.p.A.
Talks & Awards
Talks 9 talks · 5 countries
SurvKAN: A Fully Parametric Survival Model Based on Kolmogorov-Arnold Networks
International Joint Conference on Neural Networks (IJCNN)
Deep Variational Contrastive Learning for Joint Risk Stratification and Time-to-Event Estimation
International Joint Conference on Neural Networks (IJCNN)
FPBoost: Fully Parametric Gradient Boosting for Survival Analysis
European Conference on Artificial Intelligence (ECAI)
Discriminative adversarial privacy: balancing accuracy and membership privacy in neural networks
British Machine Vision Conference (BMVC)
Federated Survival Forests
International Joint Conference on Neural Networks (IJCNN)
Deep Survival Analysis for Healthcare: An Empirical Study on Post-Processing Techniques
AIxIA Workshop on Artificial Intelligence For Healthcare
Federated Survival Analysis
AIxIA Workshop on Machine Learning and Data Mining
Federated and Privacy-Preserving Learning
PhD Lecture, Politecnico di Milano
Neural Weighted A*: Learning Graph Costs and Heuristics with Differentiable Anytime A*
Conference on Machine Learning, Optimization, and Data Science (LOD)