Variable Importance Using Decision Trees
Abstract
Decision trees and random forests are well established models that not only offer good predictive performance, but also provide rich feature importance information. While practitioners often employ variable importance methods that rely on this impurity-based information, these methods remain poorly characterized from a theoretical perspective. We provide novel insights into the performance of these methods by deriving finite sample performance guarantees in a high-dimensional setting under various modeling assumptions. We further demonstrate the effectiveness of these impurity-based methods via an extensive set of simulations.
Cite
Text
Kazemitabar et al. "Variable Importance Using Decision Trees." Neural Information Processing Systems, 2017.Markdown
[Kazemitabar et al. "Variable Importance Using Decision Trees." Neural Information Processing Systems, 2017.](https://mlanthology.org/neurips/2017/kazemitabar2017neurips-variable/)BibTeX
@inproceedings{kazemitabar2017neurips-variable,
title = {{Variable Importance Using Decision Trees}},
author = {Kazemitabar, Jalil and Amini, Arash and Bloniarz, Adam and Talwalkar, Ameet S},
booktitle = {Neural Information Processing Systems},
year = {2017},
pages = {426-435},
url = {https://mlanthology.org/neurips/2017/kazemitabar2017neurips-variable/}
}