Laura Fuentes Vicente

Laura Fuentes Vicente

Ph.D candidate at INRIA-IDESP & MAP5

Talks and presentations

Talk JdS 2025

Published in Journées de la Statistique, 2025

I had the pleasure of presenting at the JdS 2025 conference in Marseille. During my talk, I shared some ongoing work on policy learning aimed at maximizing a primary outcome while controlling for adverse events. I began by introducing both the causal inference framework and the classical policy learning approach. I then motivated our choice of the EP-learning method, to which we added a constraint specifically designed to handle adverse outcomes. Finally, I presented some preliminary results based on the uncorrected version of our method, which helped highlight the need for incorporating estimation corrections in future work.

Talk Premedical team

Published in Antenne INRIA Montpellier, 2024

In this talk, I introduced the foundations of policy learning, emphasizing how data-driven approaches can enhance decision-making processes. I also provided an overview of causal inference, covering its mathematical framework and key measures of causal effect, laying the groundwork for policy learning. We then explored various techniques for policy evaluation and optimization. Lastly, I presented some initial results, with further improvements expected in the coming months.