Gaussian Processes in Machine Learning

By Carl E. Rasmussen. 2003

2,210 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Control and Systems Engineering, Probabilistic & Bayesian Methods

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Theory and Mathematics

In Control and Systems Engineering

In Probabilistic & Bayesian Methods

Citations in the corpus

Reconciling modern machine-learning practice and the classical bias–variance trade-off (2018, neutral): “Indeed, the minimum norm interpolating classifier, hn,∞, can be obtained directly by explicit norm minimization (solving an explicit system of linear equations), through SGD, or by averaging trajectories of Gaussian processes [computing the posterior mean (29)].”
Neural-net-induced Gaussian process regression for function approximation and PDE solution (2018, positive): “3 shows the approximation by the GPs with two commonly used kernels: squared exponential (SE) and Matern [12].”
Kernelized movement primitives (2017, neutral): “In addition to the factor ‘ c ’, the covariance formula in (26) differs from the covariances deﬁned in GPR and CrKR in two essential aspects.”
Gaussian Processes for Object Categorization (2010, positive): “Some popular GP models include GP classification and regression (Rasmusen and Williams 2006), non-linear dimensionality reduction using the Gaussian Process Latent Variable Model (GPLVM) (Lawrence 2004), and Mixture of GPs (Tresp 2000).”
Convex Optimization: Algorithms and Complexity (2015, neutral): “For more details on GPs and GP regression, see Rasmussen and Williams [2006].”
Learning Camera Viewpoint Using CNN to Improve 3D Body Pose Estimation (2016, neutral): “noted by rand the corresponding 3D pose by x. Training inputs and outputs are represented by R = (r1,r2,...,r N) and X = (x1,x2,...,x N) respectively, where each one was modeled by a Gaussian process [30]. Therefore the joint distribution of test input, r, and training inputs, R, is given by, N R 0”
Combining active learning and reactive control for robot grasping (2010, positive): “We propose a samplebased reinforcement learner that models the distribution of expected rewards over the continuous space of actions using Gaussian process regression (GPR) [32].”
A mobile robotic chemist (2020, positive): “Given data D and a likelihood model D ϕ θ ( | ) likelihood , this yields a posterior distribution of models using Bayes’ theorem: posterior likelihood prior The Gaussian process prior used a Matern similarity kernel, constant scaling and homoscedastic noise 41 .”
Detecting and quantifying causal associations in large nonlinear time series datasets (2017, positive): “GPDC is based on Gaussian process regression (42) and a distance correlation (43) test on the residuals, which is suitable for a large class of nonlinear dependencies with additive noise.”
Autonomous navigation of stratospheric balloons using reinforcement learning (2020, positive): “To extrapolate to other positions, we use a Gaussian process 37 to blend the balloon’s measurements with forecasts from the European Centre for Medium-Range Weather Forecast (ECMWF), using the wind forecast as a prior mean.”