Boris Baros

Boris Baros is a Research Fellow in the Department of MATH at the University of Oxford. His research focuses on optimising air-traffic control in NATS transatlantic operations, specifically flight-path planning. Funded by the National Air Traffic Service, his work addresses high-dimensional spatiotemporal stochastic control problems, with an emphasis on mean-field analysis of neural network parametrised feedback controls for entropy-regularised empirical loss minimisation. His research aims to provide theoretical guarantees for practical applications in air traffic control. Baros is also involved in teaching, offering courses such as Information Theory and Advanced Monte Carlo Methods.

Boris Baros is a PhD student at Oxford University, focusing on Stochastic Control and Machine Learning. His research explores the intersection of these fields, particularly in developing methods for learning optimal control strategies in stochastic environments. His work includes the study of generalisation bounds for learning controls, which is critical for ensuring robustness and performance in uncertain settings. Baros's research contributes to the theoretical foundations of machine learning and control theory, with applications in complex systems where uncertainty is prevalent. His academic focus reflects a commitment to advancing methodologies that bridge probabilistic models with control algorithms.