Skip to content

gp.md

Latest commit

 

History

History
78 lines (60 loc) · 1.63 KB

gp.md

File metadata and controls

78 lines (60 loc) · 1.63 KB

Gaussian Process Library

Pages = ["gp.md"]

This section describes a library for Gaussian process time series models. A technical overview of key concepts can be found in the following references.

Roberts S, Osborne M, Ebden M, Reece S, Gibson N, Aigrain S. 2013. Gaussian processes for time-series modelling. Phil Trans R Soc A 371: 20110550. http://dx.doi.org/10.1098/rsta.2011.0550

Rasmussen C, Williams C. 2006. Gaussian Processes for Machine Learning. MIT Press, Cambridge, MA. http://gaussianprocess.org/gpml/chapters/

AutoGP.GP

[Covariance Kernels](@id gp_cov_kernel)

AutoGP.GP.Node
AutoGP.GP.LeafNode
AutoGP.GP.BinaryOpNode
AutoGP.GP.pretty
AutoGP.GP.size
AutoGP.GP.eval_cov
AutoGP.GP.compute_cov_matrix
AutoGP.GP.compute_cov_matrix_vectorized
AutoGP.GP.extract_kernel
AutoGP.GP.reparameterize
AutoGP.GP.rescale

[Primitive Kernels](@id gp_cov_kernel_prim)

Notation. In this section, generic parameters (e.g., \theta, \theta_1, \theta_2), are used to denote fieldnames of the corresponding Julia structs in the same order as they appear in the constructors.

AutoGP.GP.WhiteNoise
AutoGP.GP.Constant
AutoGP.GP.Linear
AutoGP.GP.SquaredExponential
AutoGP.GP.GammaExponential
AutoGP.GP.Periodic

[Composite Kernels](@id gp_cov_kernel_comp)

AutoGP.GP.Times
AutoGP.GP.Plus
AutoGP.GP.ChangePoint

Prediction Utilities

AutoGP.Distributions.MvNormal
AutoGP.Distributions.quantile

Prior Configuration

AutoGP.GP.GPConfig