From 0b26c04f3546108779f0a493d8ef198c83b5873b Mon Sep 17 00:00:00 2001 From: Sailesh Choyal Date: Wed, 19 Jul 2017 16:46:46 +0530 Subject: [PATCH 1/2] Add links for [RW2006] --- doc/modules/gaussian_process.rst | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/doc/modules/gaussian_process.rst b/doc/modules/gaussian_process.rst index 7fae49349f342..c8da7949f4f08 100644 --- a/doc/modules/gaussian_process.rst +++ b/doc/modules/gaussian_process.rst @@ -65,7 +65,7 @@ alternative to specifying the noise level explicitly is to include a WhiteKernel component into the kernel, which can estimate the global noise level from the data (see example below). -The implementation is based on Algorithm 2.1 of [RW2006]_. In addition to +The implementation is based on Algorithm 2.1 of [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf). In addition to the API of standard scikit-learn estimators, GaussianProcessRegressor: * allows prediction without prior fitting (based on the GP prior) @@ -164,7 +164,7 @@ than just predicting the mean. GPR on Mauna Loa CO2 data ------------------------- -This example is based on Section 5.4.3 of [RW2006]_. +This example is based on Section 5.4.3 of [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf). It illustrates an example of complex kernel engineering and hyperparameter optimization using gradient ascent on the log-marginal-likelihood. The data consists of the monthly average atmospheric @@ -190,7 +190,7 @@ different properties of the signal: - smaller, medium term irregularities are to be explained by a RationalQuadratic kernel component, whose length-scale and alpha parameter, which determines the diffuseness of the length-scales, are to be determined. - According to [RW2006]_, these irregularities can better be explained by + According to [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf), these irregularities can better be explained by a RationalQuadratic than an RBF kernel component, probably because it can accommodate several length-scales. @@ -249,7 +249,7 @@ inappropriate for discrete class labels. Rather, a non-Gaussian likelihood corresponding to the logistic link function (logit) is used. GaussianProcessClassifier approximates the non-Gaussian posterior with a Gaussian based on the Laplace approximation. More details can be found in -Chapter 3 of [RW2006]_. +Chapter 3 of [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf). The GP prior mean is assumed to be zero. The prior's covariance is specified by a passing a :ref:`kernel ` object. The @@ -364,7 +364,7 @@ translations in the input space, while non-stationary kernels depend also on the specific values of the datapoints. Stationary kernels can further be subdivided into isotropic and anisotropic kernels, where isotropic kernels are also invariant to rotations in the input space. For more details, we refer to -Chapter 4 of [RW2006]_. +Chapter 4 of [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf). Gaussian Process Kernel API --------------------------- @@ -538,7 +538,7 @@ the following figure: :target: ../auto_examples/gaussian_process/plot_gpr_prior_posterior.html :align: center -See [RW2006]_, pp84 for further details regarding the +See [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf), pp84 for further details regarding the different variants of the Matérn kernel. Rational quadratic kernel From 923e0a468ccbbd0f86f2b2f3b0fd36499430c0d6 Mon Sep 17 00:00:00 2001 From: Sailesh Choyal Date: Wed, 19 Jul 2017 18:26:16 +0530 Subject: [PATCH 2/2] Update [RW2006] reference link --- doc/modules/gaussian_process.rst | 19 +++++++------------ 1 file changed, 7 insertions(+), 12 deletions(-) diff --git a/doc/modules/gaussian_process.rst b/doc/modules/gaussian_process.rst index c8da7949f4f08..94cca8999e489 100644 --- a/doc/modules/gaussian_process.rst +++ b/doc/modules/gaussian_process.rst @@ -65,7 +65,7 @@ alternative to specifying the noise level explicitly is to include a WhiteKernel component into the kernel, which can estimate the global noise level from the data (see example below). -The implementation is based on Algorithm 2.1 of [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf). In addition to +The implementation is based on Algorithm 2.1 of [RW2006]_. In addition to the API of standard scikit-learn estimators, GaussianProcessRegressor: * allows prediction without prior fitting (based on the GP prior) @@ -164,7 +164,7 @@ than just predicting the mean. GPR on Mauna Loa CO2 data ------------------------- -This example is based on Section 5.4.3 of [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf). +This example is based on Section 5.4.3 of [RW2006]_. It illustrates an example of complex kernel engineering and hyperparameter optimization using gradient ascent on the log-marginal-likelihood. The data consists of the monthly average atmospheric @@ -190,7 +190,7 @@ different properties of the signal: - smaller, medium term irregularities are to be explained by a RationalQuadratic kernel component, whose length-scale and alpha parameter, which determines the diffuseness of the length-scales, are to be determined. - According to [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf), these irregularities can better be explained by + According to [RW2006]_, these irregularities can better be explained by a RationalQuadratic than an RBF kernel component, probably because it can accommodate several length-scales. @@ -249,7 +249,7 @@ inappropriate for discrete class labels. Rather, a non-Gaussian likelihood corresponding to the logistic link function (logit) is used. GaussianProcessClassifier approximates the non-Gaussian posterior with a Gaussian based on the Laplace approximation. More details can be found in -Chapter 3 of [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf). +Chapter 3 of [RW2006]_. The GP prior mean is assumed to be zero. The prior's covariance is specified by a passing a :ref:`kernel ` object. The @@ -364,7 +364,7 @@ translations in the input space, while non-stationary kernels depend also on the specific values of the datapoints. Stationary kernels can further be subdivided into isotropic and anisotropic kernels, where isotropic kernels are also invariant to rotations in the input space. For more details, we refer to -Chapter 4 of [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf). +Chapter 4 of [RW2006]_. Gaussian Process Kernel API --------------------------- @@ -538,7 +538,7 @@ the following figure: :target: ../auto_examples/gaussian_process/plot_gpr_prior_posterior.html :align: center -See [RW2006](http://www.gaussianprocess.org/gpml/chapters/RW.pdf), pp84 for further details regarding the +See [RW2006]_, pp84 for further details regarding the different variants of the Matérn kernel. Rational quadratic kernel @@ -601,12 +601,7 @@ shown in the following figure: References ---------- - * `[RW2006] - `_ - **Gaussian Processes for Machine Learning**, - Carl Eduard Rasmussen and Christopher K.I. Williams, MIT Press 2006. - Link to an official complete PDF version of the book - `here `_ . +.. [RW2006] Carl Eduard Rasmussen and Christopher K.I. Williams, "Gaussian Processes for Machine Learning", MIT Press 2006, Link to an official complete PDF version of the book `here `_ . .. currentmodule:: sklearn.gaussian_process