Gaussian Process Regression
Gaussian Process Regression (GPR) is a Bayesian non-parametric method used for regression tasks, aiming to predict a continuous output variable based on input data while providing uncertainty estimates. Current research emphasizes improving GPR's scalability and robustness, focusing on techniques like dividing local Gaussian processes for continual learning, tensor network methods for high-dimensional data, and efficient kernel selection and subsampling strategies. These advancements enhance GPR's applicability across diverse fields, including system identification, time series forecasting, safety-critical control systems, and scientific modeling where accurate uncertainty quantification is crucial.
Papers
Decentralized Event-Triggered Online Learning for Safe Consensus of Multi-Agent Systems with Gaussian Process Regression
Xiaobing Dai, Zewen Yang, Mengtian Xu, Fangzhou Liu, Georges Hattab, Sandra Hirche
Whom to Trust? Elective Learning for Distributed Gaussian Process Regression
Zewen Yang, Xiaobing Dai, Akshat Dubey, Sandra Hirche, Georges Hattab
Kernel-, mean- and noise-marginalised Gaussian processes for exoplanet transits and $H_0$ inference
Namu Kroupa, David Yallup, Will Handley, Michael Hobson
Imitation learning for sim-to-real transfer of robotic cutting policies based on residual Gaussian process disturbance force model
Jamie Hathaway, Rustam Stolkin, Alireza Rastegarpanah
Projecting basis functions with tensor networks for Gaussian process regression
Clara Menzen, Eva Memmel, Kim Batselier, Manon Kok
Stochastic Gradient Descent for Gaussian Processes Done Right
Jihao Andreas Lin, Shreyas Padhy, Javier Antorán, Austin Tripp, Alexander Terenin, Csaba Szepesvári, José Miguel Hernández-Lobato, David Janz