Caterina De Bacco

Caterina De Bacco

Associate Professor

TU Delft

Physics for Inference and Optimization

I like to approach research by addressing application-oriented problems involving domain experts from different disciplines via developing models and algorithms derived from statistical physics principles.

From September 2024 I am Associate Professor at TU Delft, part of the NAS group, in Delft, Netherlands.

Between July 2018 and August 2024 I was an Independent Research Group Leader at Cyber Valley, Max Planck Institute for Intelligent Systems in Tübingen, Germany.

I obtained a degree in Physics at University of Padova and a PhD in Statistical Physics at Universitè Paris Sud 11 in 2015, advised by Silvio Franz and Satya Majumdar with a Marie Curie ITN fellowship .
After that, I was a Program Postdoc at the Santa Fe Institute working with Cris Moore and then at the Columbia Data Science Institute working with David Blei.

Interests

  • Inference on networks
  • Routing Optimization
  • Probabilistic Modeling
  • Statistical physics

Education

  • PhD in Statistical Physics, 2015

    Universitè Paris Sud 11

  • MSc in Theoretical Physics, 2012

    Università di Padova

  • BSc in Physics, 2010

    Università di Padova

News

  • We are hiring PhD students!!

This is part of NExTWORKx, a strategic partnership between the Telecom and ICT service provider KPN and Delft University of Technology.

Deadline for application is December 1st 2024.

Do you have any interesting work addressing any aspect of statistical inference and machine learning for complex network data and models? Then please submit your article here.

Deadline for submissions is May 31st 2025 (flexible).

Recent Publications

Andrea Della Vecchia, Kibidi Neocosmos, Daniel B. Larremore, Moore Cristopher, Caterina De Bacco. Model for efficient dynamical ranking in networks. Phys. Rev. E 110, 034310, 2024.

Preprint PDF Code Source Document

Daniela Leite, Caterina De Bacco. Similarity and economy of scale in urban transportation networks and optimal transport-based infrastructures. Nat Commun 15, 7981 (2024), 2024.

Preprint PDF Code Source Document

Anna Badalyan, Nicolò Ruggeri, Caterina De Bacco. Structure and inference in hypergraphs with node attributes. Nat Commun 15, 7073 (2024), 2024.

Preprint PDF Code Source Document

Abdullahi Adinoyi Ibrahim, M Muehlebach, Caterina De Bacco. Optimal Transport with Constraints: From Mirror Descent to Classical Mechanics. Phys. Rev. Lett. 133, 057401, 2024.

Preprint PDF Code Source Document

Martina Contisciani, M Hobbhahn, E a Power, P Hennig, Caterina De Bacco. Flexible inference in heterogeneous and attributed multilayer networks. Submitted, 2024.

Preprint PDF Code

Hadiseh Safdari, Caterina De Bacco. Community detection and anomaly prediction in dynamic networks. Submitted, 2024.

Preprint PDF

See all publications

Research

I study large interacting systems following two main research directions to tackle the problem under different angles.

Inference on networks

Complex interacting systems are often represented by large datasets containing a considerable amount of information. The question is how to capture the relevant macroscopic behavior by retaining only a small amount of information. This can be formalized as an inference problem where we want to extract a set of model parameters capable of describing the observed system and making future predictions.
In particular, I am interested in analyzing systems where elements interact in multiple ways, as in multilayer networks, and systems displaying hidden hierarchies that might play a role in determining the interaction patterns that we observe.

Routing Optimization on networks

Optimizing traffic on a network is a relevant problem in situations where traffic congestion has a big impact on transmission performance. This can be formalized as a computationally-hard constrained optimization problem where interactions are non-local and a global optimization is required.
I investigate this problem by adopting approaches that combine insights from statistical physics and a methodologies developed in optimal transport theory.

The Team

Past Members

Avatar

Daniela Leite

PhD student

Avatar

Kibidi Neocosmos

Research Intern

Avatar

Anna Badalyan

Research Intern

Avatar

Nicolò Ruggeri

PhD student

Avatar

Abdullahi Adinoyi Ibrahim

PhD student

Avatar

Martina Contisciani

PhD student

Avatar

Hadiseh Safdari

Postdoctoral researcher

Avatar

Alessandro Lonardi

PhD student

Avatar

Diego Baptista Theuerkauf

PhD student

Avatar

Laura Iacovissi

Research Intern

Avatar

Nicolò Zottino

PhD student

Alumni

Andrea Della Vecchia

Intern, Master Thesis, 2019

Johannes Schulz

Intern, Master thesis, 2022

Sameh Othman

Research Intern, 2021-22

Aneesh Barthakur

Research Intern, 2022-23

Emanuele Pigani

Intern, Master Thesis, 2019

Lorenzo Ferretti

Intern, Master Thesis, 2019-20

Tainá Turella Caetano dos Santos

Research Intern, 2023

Teaching

  • Performance Analysis: for the MSc Electrical Engineering at TU Delft, IN4-341, Q2 2024-2025.
  • Advanced Probabilistic methods for Machine Learning and Applications: for the Master Degree in Machine Learning at University of Tübingen, 2020-21 and 2021-22.
  • Advanced Probabilistic methods for Machine Learning and Applications: class taught together with Prof. Isabel Valera for the Master Degree in Machine Learning at University of Tübingen, 2019-20.

Contact