Advances in Scalable Gaussian Processes

José Miguel Hernández Lobato

  Universidad de Cambridge

Resumen/Abstract

Gaussian processes are a powerful tool for uncertainty quantification and decision-making, but their scalability is challenged by cubic computational costs. This talk explores key advancements in scalable Gaussian processes that leverage stochastic optimization through iterative solvers. First, I will introduce a simple and effective stochastic optimization algorithm, demonstrating its superiority over preconditioned conjugate gradients and variational approximations in regression and Bayesian optimization tasks. Second, I will present innovative strategies for hyperparameter optimization in large datasets, including pathwise gradient estimators, warm-starting solvers, and early stopping, achieving significant speed-ups and improved efficiency. Together, these contributions push the boundaries of Gaussian process scalability, enabling their use in large-scale and ill-conditioned settings.

Profesor EPS Proponente: Daniel Hernández.

Curriculum ponente

Jose Miguel Hernandez Lobato is a lecturer in Machine Learning at the Department of Engineering of the University of Cambridge. Before that, he was a postdoctoral fellow at the Harvard School of Applied Sciences and Engineering from Sept. 2014 to Sept. 2016. His research interests are in Bayesian optimization, scalable methods for approximate inference, and flexible probabilistic modeling of data. Jose Miguel's research is driven by machine learning applications to expensive optimal design problems in engineering. Before joining Harvard, Jose Miguel was a postdoctoral research associate at the Department of Engineering of the University of Cambridge where he worked in a collaboration project with the Indian multinational company Infosys Technologies. From December 2010 to May 2011, Jose Miguel was a teaching assistant at the Computer Science Department at Universidad Autónoma de Madrid (Spain), where he obtained his Ph.D. and M.Phil. in Computer Science in December 2010 and June 2007, respectively. Jose Miguel also received a B.Sc. in Computer Science from this institution in June 2004, with a special prize for the best academic record on graduation.