GPAR.jl: A simple package for autoregressive gaussian process functionals in Julia

A Julia package for modeling hysteretic data with Gaussian process regression!

What is an autoregressive Guassian Process functional?

This package is built using KernelAbstractions.jl to enable all models to run on parallel hardware (GPU or CPUs). My intention is not to replace existing GP packages such as AbstractGPs.jl or GaussianProcesses.jl. Rather, I have a very specific usecase that requires a special kind of gaussian process and I need it to run on GPU. These packages are not compatible with GPU hardware, so this is my best crack at an implementation.