PhD Thesis by Omar Fabian González

We are pleased to announce the public defense of the doctoral thesis by Omar Fabian González, entitled “Bayesian Computational Inference in High-Dimensional Systems”, conducted under the supervision of Professor Joaquín Míguez Arenas, which took place on Wednesday, April 8th at 11:00 AM in the Aula de Grados (Adoración de Miguel) at Universidad Carlos III de Madrid. The thesis addresses key challenges in stochastic filtering, a fundamental area of applied probability with applications in communications, artificial intelligence, finance, and physics, focusing in particular on the curse of dimensionality and model misspecification. It provides a novel theoretical analysis of nested importance sampling methods, showing that, under suitable conditions, approximation errors scale only polynomially with the dimension—and may even remain uniformly bounded—offering new insights into the scalability of these techniques. In addition, the dissertation develops a principled interpretation of nudging methods as modifications of state-space model dynamics that improve marginal likelihood, alongside an operator-theoretic framework that unifies stability results and clarifies the propagation of modeling errors. Altogether, this work represents a significant contribution to the theoretical understanding and practical robustness of Bayesian filtering in high-dimensional settings.

MLCARE Kickoff Meeting

MLCARE (Machine Learning Computational Advancements for peRsonalized mEdicine, HORIZON–MSCA–2024–DN–101226456) is a Marie Skłodowska-Curie Doctoral Network focused on advancing AI methods for precision medicine. The project aims to unlock the largely untapped potential of biomedical data—ranging from genomics and imaging to electronic health records—by developing trustworthy and explainable computational models capable of integrating multimodal patient information. Over the next four years, MLCARE will also train 14 doctoral candidates at the intersection of machine learning, computational biology, and healthcare, contributing to Europe’s leadership in data-driven, patient-centred medicine.

On 11–12 February 2026, the network officially launched with a two-day Kick-off Meeting at UC3M’s Puerta de Toledo Campus. The first day was devoted to internal consortium discussions on governance, recruitment, and coordination of research activities, establishing a common roadmap for the project. The second day featured an Open Day open to the wider community, with invited talks by speakers from AstraZeneca Spain, the Max Planck Institute of Biochemistry, the Pasteur Institute, and Novo Nordisk. The sessions addressed topics spanning AI in metabolic health and intensive care, topological learning, and statistical genomics, offering an overview of current challenges and opportunities in computational biomedicine while introducing MLCARE to prospective applicants and collaborators.