基于随机森林和增强森林的区间预测 R语言版

[](https://api.travis-ci.com/calakus/RFpredInterval.svg?branch=main) # RFpredInterval R package which implements Prediction Intervals with Random Forests and Boosted Forests `RFpredInterval` package has two main R functions as below: * `pibf`: Constructs prediction intervals with **P**rediction **I**ntervals with **B**oosted **F**orests (PIBF) proposed by Alakus et al. (2021) * `rfpi`: Constructs prediction intervals with 15 distinct variations to produce **P**rediction **I**ntervals with **R**andom **F**orests (RFPI) proposed by Roy and Larocque (2020) ## Installation This package is available on [CRAN](https://CRAN.R-project.org/package=RFpredInterval). Alternatively, you can install **RFpredInterval** from GitHub using the `devtools` package. Run the following code in `R` to install: ```R if (!require(devtools)) { install.packages("devtools") library(devtools) } devtools::install_github('calakus/RFpredInterval') ``` ## References - Alakus, C., Larocque, D., and Labbe, A. (2021). RFpredInterval: An R Package for Prediction Intervals with Random Forests and Boosted Forests. *arXiv preprint* <https://arxiv.org/abs/2106.08217>. - M.-H. Roy and D. Larocque. Prediction intervals with random forests. *Statistical Methods in Medical Research*, 29(1):205–229, Jan. 2020. URL <https://doi.org/10.1177/0962280219829885>.