grf: Generalized Random Forests
Julie Tibshirani, Susan Carleton Athey, Erik Sverdrup, Stefan Wager · 2017
Forest-based statistical estimation and inference. GRF provides non-parametric methods for heterogeneous treatment effects estimation (optionally using right-censored outcomes, multiple treatment arms or outcomes, or instrumental variables), as well as least-squares regression, quantile regression, and survival regression, all with support for missing covariates.