Summary of Accurate Estimation Of Feature Importance Faithfulness For Tree Models, by Mateusz Gajewski et al.
Accurate estimation of feature importance faithfulness for tree models
by Mateusz Gajewski, Adam Karczmarz, Mateusz Rapicki, Piotr Sankowski
First submitted to arxiv on: 4 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper proposes a novel metric called PGI squared to evaluate the predictive faithfulness of feature rankings (or attributions) for decision tree-based regression models. The metric can be computed efficiently and accurately for arbitrary independent feature perturbation distributions, unlike traditional Monte Carlo sampling methods that are prone to inaccuracies. The authors also introduce a method for ranking features based on their importance for a model’s predictions using PGI squared. Experimental results show that this approach can identify globally important features better than the state-of-the-art SHAP explainer in some cases. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how well feature rankings (or attributions) match up with what they actually do to help predict things. They came up with a new way to measure this, called PGI squared, that works really well for certain kinds of models and can even identify the most important features. This is useful because it helps us understand which features are most important in making predictions. |
Keywords
* Artificial intelligence * Decision tree * Regression
