Eduardo did his undergraduate studies in Physics and obtained a master’s degree in applied mathematics from the University of Waterloo in Canada. After a career in secondary mathematics education in an international setting, he obtained a Master's degree in Machine Learning and Data Mining from the Jean Monnet University in Saint Étienne. He is currently a second-year PhD student at the Hubert Curien Laboratory, where he is working on integrating physical knowledge in Machine Learning.


The integration of background physical knowledge in machine learning (ML) algorithms is a promising emerging topic, which has recently produced interesting results in both ML and Physics.
In ML, one typically integrates contextual knowledge in situations where data is insufficient to build sophisticated models but a good physical model is available. In Physics, one would like to leverage physics-agnostic ML models to make use of abundant available data to validate or refine existing physical models -- or even propose new ones. But in many cases, we are far from the high physical knowledge/high experimental data regime, as one has only a partial or incomplete physical model and few experimental data -- for example, when studying a complex physical situation in which data is difficult or costly to acquire.
I am interested in the few-data/partial-knowledge regime, and on developing ML techniques and tools that can be used in this common situation.