Benjamin Lu
2020-Jan-20 18:10 UTC
[R-pkgs] forestError 0.1.0: Random forest conditional prediction intervals, squared errors, and biases
Hi all, I?m writing to introduce a new package, forestError. This package estimates conditional response quantiles (prediction intervals), conditional mean squared prediction errors, conditional biases, and conditional prediction error distribution functions for random forests using new methods proposed in Lu and Hardin (2019+) <arXiv:1912.07435>. The package does not grow specially designed random forests, but rather sits on top of the standard random forest algorithm. In particular, the package can estimate the above quantities for regression random forests built using the randomForest, randomForestSRC, ranger, and quantregForest packages. I hope that you find this useful. Please feel free to reach out with feedback or questions. CRAN: https://cran.r-project.org/package=forestError Github: https://github.com/benjilu/forestError Reference: https://arxiv.org/abs/1912.07435 Best, Benji Lu [[alternative HTML version deleted]]