Cláudia Soares
I am a researcher at NOVA LINCS and an Assistant Professor at the Department of Computer Science of the NOVA School of Science and Technology. My work is broadly on trustworthy ML for critical Systems, like healthcare and physical systems (swarms of artificial agents, underwater unmanned vehicles, and satellites).
My Erdös number is 4.
If you would like to join our group as a Postdoctoral Researcher, Ph.D. student, M.Sc., or B.Sc. student, see this page.
Selected Recent Papers
- ASTRA: ADMM-Accelerated Topology Reconfiguration for Dynamic Satellite Constellations
A method for dynamic satellite constellation reconfiguration with regret guarantees and minimal pre-existing knowledge.
- CLASP: Online learning algorithms for Convex Losses And Squared Penalties
Online learning algorithms for convex losses and squared penalties.
- A Multi-Token Coordinate Descent Method for Semi-Decentralized Vertical Federated Learning
A semi-decentralized vertical federated learning method based on multi-token coordinate descent.
- OrbitZoo: Real Orbital Systems Challenges for Reinforcement Learning
Challenges for reinforcement learning in real orbital systems.
Some Upcoming events
- SPAICE Conference. AI in and for Space
I will be chairing a session and my group will be presenting our work on AI for space applications at the SPAICE conference. We will be delighted to Connect.
- NeurIPS 2026
João Norberto will present our work in Online Learning with guarantees even under imperfect updates, applied to the combinatorial problem of connecting satellites in large constellations, at the NeurIPS 2026 main conference. Drop by our poster on Dec 7 at 10 pm PST!