Fill in some content

Author: theogf

abstractgps.md

@@ -8,5 +8,8 @@
 @def github_repo = "JuliaGaussianProcesses/AbstractGPs.jl"
 
 \begin{section}{name="AbstractGPs.jl", title="AbstractGPs"}
-
+\lead{
+AbstractGPs give you all the tools to build GP models from scratch.
+Check out the tutorials, the docs and the github repo for more details!
+}
 \end{section}

gplikelihoods.md

@@ -9,4 +9,24 @@
 
 \begin{section}{name="GPLikelihoods.jl", title="GPLikelihoods.jl"}
 
+\lead{
+GPLikelihoods aims at facilitating the use of latent models with non-conjugate likelihoods.
+The goal is to represent different models such as classification $y \sim \textrm{Bernoulli}(g(f(x))))$ or event counting ${y \sim \textrm{Poisson}(g(f(x)))}$ given an adapted transformation $g$.
+
+**Note that this package is still a WIP**
+}
+
+\end{section}
+
+\begin{section}{title="Current implementations"}
+
+Given $f$ a latent GP, or $\{f_i\}$ a collection of latent GPs we have the following implementation
+
+@@list
+- Gaussian likelihood : $y \sim f + \epsilon$ where $\epsilon \sim \mathcal{N}(0, \sigma)$
+- Bernoulli likelihood : $y \sim \textrm{Bernoulli}(p)$ where $p = \sigma(f)$
+- Poisson likleihood : $y \sim \textrm{Poisson}(\lambda)$ where $\lambda = \exp(f)$
+- Categorical likelihood : $y \sim \textrm{Categorical}(\pi)$ where $\pi = \textrm{softmax}(\{f_i\})$
+@@
+
 \end{section}

index.md

@@ -6,6 +6,10 @@
 
 \lead{This is the website to host the general documentation for the packages [KernelFunctions.jl](https://github.com/JuliaGaussianProcesses/KernelFunctions.jl), [AbstractGPs.jl](https://github.com/JuliaGaussianProcesses/AbstractGPs.jl), [GPLikelihoods.jl](https://github.com/JuliaGaussianProcesses/GPLikelihoods.jl) and more to come!}
 
+\lead{
+**Note that this website is still a work in progress and a lot of things are missing/buggy, for now check out the docs of each package, [submit an issue](https://github.com/JuliaGaussianProcesses/JuliaGaussianProcesses.github.io/issues/new) or don't hesitate to participate by [creating a pull request](https://github.com/JuliaGaussianProcesses/JuliaGaussianProcesses.github.io/pulls?q=is%3Apr+is%3Aopen+sort%3Aupdated-desc)**
+}
+
 \lead{You will find in this website a collection of [tutorials](tutorials) combining one or more packages from the JuliaGPs ecosystem for typical machine learning or statistics problems}
 
 \lead{For each of the packages we also provide ressources and documentation to get a better understanding of the tools and the topics you can access them here: