Sites Inria

Version française

Prize and distinction


Marco Lorenzi, ANR Young Researchers (JCJC) laureate

© Inria G. Scagnelli

The JCJC instrument allows project leaders to work independently on a specific scientific theme. It promotes responsibility and the capacity for scientific innovation.

Marco Lorenzi, researcher in the Epione team at the Inria Sophia Antipolis - Méditerranée centre, is the new recipient of the ANR JCJC (Young researchers) grant.

You have received an ANR Young Researchers award. Can you tell us about the program?

The purpose of the ANR Young Researchers Initiative is to support young researchers in the development of their research projects. To be eligible for it, you must have defended your thesis for less than ten years. This support consists of funding to help the candidate develop his or her research project independently. Thanks to these financial resources, it is possible to envisage the creation of a first research team to work on a chosen topic, as well as the development of international collaborations and the organization of scientific events. 

Which research project did you present?

My research project involves the development of deep learning methods to study clinical data hosted in different hospitals, while respecting patients' privacy.  The application of statistical learning to biomedical data imposes drastic constraints: models must respect the anonymity and non-transferability of information from one centre to another, while taking into account the enormous size and variability of the data. Together with my collaborators, we will take up this challenge by reformulating Bayesian non-parametric approaches in the field of "federated learning "*. In this context, statistical learning on complex and heterogeneous data but only with parameter distributions. By analyzing shared data from several hospitals, we will be able to study the effects of genetic variants in the development of neurodegenerative diseases, such as Alzheimer's disease, and we will develop methods to predict sudden death from a network of several clinical sites.

How will you benefit from the award?

Thanks to the ANR JCJC funding I will be able to create a small team including a doctoral student and a post-doctoral fellow to work with me on this project. Our work will be done in an international partnership, in collaboration with global experts in the fields of medical imaging, genetics, and neurology.
Clearly, this scholarship will allow me to strengthen my existing collaborations and develop new opportunities for technology transfer in the health field. This project represents a unique opportunty to acquire new scientific responsibilities and to consolidate my position in my research field. 

*Federated learning is an automatic learning technique that involves an algorithm on several peripheral devices or decentralized servers containing local data samples, without exchanging data samples

Keywords: Inria - Research Centre Sophia Antipolis - Méditerranée Deep learning Federated learning ANR JCJC