ParetoSmooth.jl is a Julia package for efficient approximate leave-one-out cross-validation for fitted Bayesian models. We compute LOO-CV using Pareto smoothed importance sampling (PSIS), a modification of importance sampling. More details can be found in Vehtari, Gelman, and Gabry (2017).
If you use this library, please remember to cite both:
@misc{ParetoSmooth.jl,
author = {Carlos Parada <cdp49@cam.ac.uk>},
title = {ParetoSmooth.jl},
url = {https://github.com/TuringLang/ParetoSmooth.jl},
version = {v0.7.1},
year = {2021},
month = {6}
}
and:
@Article{Vehtari2017,
author={Vehtari, Aki
and Gelman, Andrew
and Gabry, Jonah},
title={Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC},
journal={Statistics and Computing},
year={2017},
month={Sep},
day={01},
volume={27},
number={5},
pages={1413-1432},
issn={1573-1375},
doi={10.1007/s11222-016-9696-4},
url={https://doi.org/10.1007/s11222-016-9696-4}
}