Poster
Kernel Interpolation for Scalable Online Gaussian Processes
Samuel Stanton · Wesley Maddox · Ian Delbridge · Andrew Gordon Wilson
Keywords: [ Models and Methods ] [ Gaussian Processes ]
Abstract:
Gaussian processes (GPs) provide a gold standard for performance in online settings, such as sample-efficient control and black box optimization,
where we need to update a posterior distribution as we acquire data in a sequential online setting. However, updating a GP posterior to accommodate even a single new observation after having observed points incurs at least computations in the exact setting. We show how to use structured kernel interpolation to efficiently reuse computations for constant-time online updates with respect to the number of points , while retaining exact inference. We demonstrate the promise of our approach in a range of online regression and classification settings, Bayesian optimization, and active sampling to reduce error in malaria incidence forecasting.
Chat is not available.