Christian Bongiorno

Christian Bongiorno

Associate Professor at CentraleSupélec, Université Paris-Saclay. Laboratoire de Mathématiques et Informatique pour la Complexité et les Systèmes (MICS)

Research interests

Financial Markets and Portfolio Management

• Application of machine learning techniques to portfolio optimization.
• Neural networks for dynamic portfolio management.
• Random matrix theory applied to covariance matrices in financial systems.
• Information theory in financial markets.

Complex Systems and Networks

• Statistical characterization of complex networks.
• Dynamical systems analysis in urban mobility and transportation.

Machine Learning and Statistical Testing

• Time-irreversibility estimation with machine learning.
• Clustering algorithms
• Multivariate statistical testing.

Data Analysis and Modeling

• Quantitative investigation of real-world systems.
• Agent-based models based on empirical observations.

Bio

Dr. Christian Bongiorno earned his Ph.D. in Physics from the Università di Palermo in 2017, where he specialized in the statistical characterization and agent-based modeling of air transportation systems. He conducted postdoctoral research at the Politecnico di Torino, focusing on control-oriented models for mobility-on-demand systems. Additionally, Dr. Bongiorno was a research affiliate at the Massachusetts Institute of Technology's Senseable City Laboratory, where he explored human navigation behaviors in urban environments.

As an Associate Professor at CentraleSupélec, Université Paris-Saclay, Dr. Bongiorno leads research initiatives that bridge theoretical frameworks with practical applications in complex systems. His interdisciplinary approach emphasizes data analysis and modeling based on observable phenomena, reflecting his philosophy that physics is fundamentally the study of nature through empirical investigation. Currently, his research is centered on financial markets, particularly portfolio management, where he leverages machine learning and statistical methods to uncover hidden structures within complex financial systems.

Publications

• Vector-based Pedestrian Navigation in Cities
  Nature Computational Science
  Link

• Covariance Matrix Filtering with Bootstrapped Hierarchies
  PloS One
  Link

• Filtering time-dependent covariance matrices using time-independent eigenvalues
  Journal of Statistical Mechanics: Theory and Experiment
  Link

Source code

• Time-Irreversibility-Estimator
  Utilizes gradient boosting to measure entropy production from multivariate dynamical systems.

• BAHC
  Bootstrap Average Hierarchical Clustering – A Python package for filtering covariance matrices using hierarchical clustering and bootstrap aggregation.

• SVHC
  Statistically Validated Hierarchical Clustering – An algorithm to extract statistically significant hierarchical subpartitions from clustering results.

Internships

I am open to supervising Master theses that align with my research interests and expertise. Prospective students are encouraged to ensure their proposed topics are compatible with my work before reaching out. Please note that the administrative process for internships typically requires more than three months before commencement.