Ceci n’est pas une bio
José Manuel de Frutos — PhD researcher in Machine Learning at Universidad Carlos III de Madrid. I work on statistical machine learning, implicit generative models, rank-based methods, and divergence estimation. My research also covers distribution comparison, f-divergences, and density estimation. Former Devo (Loxcope) software engineer and ESO data engineer; CERN Openlab alumnus.
Research
I study statistical machine learning, with a particular emphasis on implicit generative models and principled distribution comparison. My work develops rank-statistic methods for approximating f-divergences and divergence estimation, robust training methods for generative models, and density estimation and density approximation techniques for neural implicit samplers. I am also interested in sliced divergences, high-dimensional distribution comparison, and their connections to optimal transport.
News
Paper accepted at ICML 2026
Our paper Approximating f-Divergences with Rank Statistics, joint work with Viktor Stein, has been accepted at ICML 2026!
Paper accepted in JMLR
Paper accepted in the Journal of Machine Learning Research (March 2026)! Robust training of implicit generative models for multivariate and heavy-tailed distributions with an invariant statistical loss
Paper accepted at AISTATS 2026
Accepted at AISTATS 2026 as a Spotlight presentation (January 2026)! Explicit Density Approximation for Neural Implicit Samplers Using a Bernstein-Based Convex Divergence. Jose Manuel de Frutos, Pablo M. Olmos, Manuel Alberto Vázquez, Joaquín Míguez
Paper accepted at AISTATS 2024
Paper accepted at AISTATS 2024 (January 2024)! Training Implicit Generative Models via an Invariant Statistical Loss. Jose Manuel de Frutos, Pablo M. Olmos, Manuel Alberto Vázquez, Joaquín Míguez