Research

In contemporary science, I find statistics to be an essential tool to understand phenomena, and to draw reliable conclusions. I am broadly interested in the analysis of large and complex data, with particular emphasis on structured objects such as functional data and networks. I am also keen on the development of new theory and methods to integrate state-of-the-art ML prediction algorithms into stable inferential frameworks. My work is often inspired by problems in causal inference or interdisciplinary collaborations with researchers in the sciences.

Some projects

These are some of the projects I have (somehow recently) been working on:

• Doubly robust estimation of functional average treatment effects, together with Tobia Boschi, Francesca Chiaromonte, Edward H. Kennedy, and Matthew Reimherr. The preprint is available here.

• Functional Feature Selection, jointly with Tobia Boschi, Francesca Chiaromonte, and Matthew Reimherr. This is a poster on FAStEN, an efficient algorithm for feature selection and coefficient estimation in functional regression models. The paper on FAStEN has been published on JCGS in November 2024, and the preprint is available here

• The Weighted Climate Dataset, together with Marco Gortan, Giorgio Fagiolo, and Francesco Lamperti. The papers introducing this tool are available here and here

• COVID-19, a joint work with Tobia Boschi, Jacopo Di Iorio, Francesca Chiaromonte, and Marzia A. Cremona. Results of our first paper appeared in the news [1], [2, Italian], [3, Italian], and as a SciPod’s audio book. A second paper is available as a preprint here!

• Venture Capital investments, an analysis of the determinants of success of start-ups financing through venture capitals. Results, which were obtained and validated with a combination of network, functional data and multivariate analysis tools, were published here (as a journal article) and here (as a conference proceeding). This is a very short and informal presentation of the results (in Italian).

You can find my publication record in my CV and on my Google Scholar profile.