Gerardo Duran-Martin
Statistical machine learning
University of Oxford | Oxford-Man Institute
I'm a postdoctoral researcher at the Oxford-Man Institute, University of Oxford. I completed my PhD in statistical machine learning at Queen Mary University of London under the supervision of Kevin P. Murphy and Alexander Y. Shestopaloff.
My research centers on developing learning algorithms for sequential decision-making and sequential (online) learning in dynamic, non-stationary environments. I follow a probabilistic approach, leveraging filtering techniques to address key challenges in online continual learning, neural contextual bandits, time-series forecasting, and reinforcement learning.
I’m particularly interested in developing adaptive agents that make decisions under uncertainty and model misspecification. My work explores how to leverage statistical tools to build agents that identify and prioritise relevant information, especially when dealing with stale data or dynamic, ever-changing environments.