timbR: Tree interpretation methods based on ranger

Lea Louisa Kronziel, Björn-Hergen Laabs

Introduction

timbR is a collection of methods for the interpretation of random forests trained by ranger.

In version 1.0 most representaive trees can be selected based on different tree based distance measures. In version 2.0 artificial representative trees can be created based on different tree based distance measures. In addition, uncertainty quantification for regression using inductive conformal prediction (ICP) and Mondrian ICP is available.

Installation

R version

To install the timbR R package, run

install.packages("devtools")
library(devtools)
devtools::install_github("imbs-hl/timbR")

If you find any bugs, or if you experience any crashes, please report to us.

Please cite our paper if you use timbR.

References

• Laabs, B.-H., Westenberger, A., & König, I. R. (2024). Identification of representative trees in random forests based on a new tree-based distance measure. Advances in Data Analysis and Classification, 18:363–380.
• Bannerjee, M., Ding, Y., Noone, A.-M. (2012) Identifying representative trees from ensembles. Stat in Med 31:1601-16.
• Laabs, B.-H., Kronziel, L. L., König, I. R., & Szymczak, S. (2024). Construction of artificial most representative trees by minimizing tree-based distance measures. In L. Longo, S. Lapuschkin, & C. Seifert (Eds.), Explainable Artificial Intelligence:290–310.