On the Robustness of Global Feature Effect Explanations

MCML Authors

Giuseppe Casalicchio

Dr.

→ Group Bernd Bischl
 Statistical Learning and Data Science

Bernd Bischl

Prof. Dr.

Director

Statistical Learning and Data Science

Abstract

We study the robustness of global post-hoc explanations for predictive models trained on tabular data. Effects of predictor features in black-box supervised learning are an essential diagnostic tool for model debugging and scientific discovery in applied sciences. However, how vulnerable they are to data and model perturbations remains an open research question. We introduce several theoretical bounds for evaluating the robustness of partial dependence plots and accumulated local effects. Our experimental results with synthetic and real-world datasets quantify the gap between the best and worst-case scenarios of (mis)interpreting machine learning predictions globally.

inproceedings BCB+24

---

ECML-PKDD 2024

European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases. Vilnius, Lithuania, Sep 09-13, 2024.

---

Authors

H. Baniecki • G. Casalicchio • B. Bischl • P. Biecek

---

Links

DOI

---

Research Area

 A1 | Statistical Foundations & Explainability