I’m a 2nd year Ph.D. student in Operations Research and Financial Engineering (ORFE) at Princeton University.
My work lies at the intersection of probabilistic machine learning and non-convex optimization.
I am fortunate to be advised by Prof. Boris Hanin.
Before my Ph.D., I obtained a Master’s degree in Mathematics (track: Stochastics and Data Science) from the University of Turin, and I was a Senior Allievo at Collegio Carlo Alberto.
My earlier work focused on Probability Theory and explored Bayesian neural networks and their infinite-width prior and posterior limits. I was fortunate to work with Prof. Stefano Favaro and Prof. Dario Trevisan.
I obtained my Bachelor’s degree in Mathematics (track: Computational Math) from the University of Pisa, with a thesis on the universality of neural networks.
- e-mail:
fcaporali [at] princeton.edu - github: caporali
- linkedin: francescocaporali
News
- Jul. 2025: Attended COLT 2025 (Lyon) to present a poster at the TASC Workshop.
- Jun. 2025-Jun. 2026: Working on the stability and implicit bias of step-size schedulers for GD, and studying principled first-order methods for convergence to flat configurations of neural networks.
- Jun. 2026-Aug. 2026: I will intern as an Applied Scientist at AWS AI Labs.
Publications / Preprints
- F. Caporali, S. Favaro, D. Trevisan, Student-$t$ processes as infinite-width limits of posterior Bayesian NNs, In: TASC Workshop, COLT 2025