Introducing our Doctoral Candidates
Today - DC04 - Richard Oliveira - CNR
What is your research about?
My research studies the optimization-theoretic foundations of machine learning in distributed settings, with emphasis on how implicit regularization emerges through networked interactions. A central focus is understanding how graph structure, consensus dynamics, and spectral properties influence the stability, convergence, and generalization of gradient-based methods.
Why is this research needed?
This research is needed to understand how regularization emerges in distributed learning dynamics and how it governs the stability, convergence, and generalization of networked learning algorithms. By clarifying the interplay between network structure, consensus processes, and implicit regularization, this work enables the principled design of scalable and reliable learning methods in localization and mapping applications.
Where does your research takes place?
Consiglio Nazionale delle Ricerche, Politecnico di Milano and KUL
Who will benefit from this research?
This research will benefit researchers and practitioners working on tomographic imaging, localization, and mapping systems that rely on distributed and data-driven inference.
When did your project start and when will it finish?
Project started on September 1st 2025 and will end on September 1st 2028
