An R Package with Boosting and SMOTEBoost implementations for Regression Tasks
Data pre-processing methods such as resampling strategies are the most common approach to tackling imbalanced domain learning tasks. This package encompasses multiple resampling-based boosting strategies for the task of extreme value prediction/imbalanced regression.
References
- Nuno Moniz, Rita Ribeiro, Vitor Cerqueira, Nitesh Chawla (2018). "SMOTEBoost for Regression: Improving the Prediction of Extreme Values", Proceedings of the 2018 IEEE 5th International Conference on Data Science and Advanced Analytics.
To install from github use the following command lines in R:
library(devtools) # You need to install this package!
install_github("nunompmoniz/ReBoost",ref="master")
After installation the package can be used loaded by doing:
library(ReBoost)