Pitfalls in machine learning interpretability: manipulating partial dependence plots to hide discrimination

DOI10.1016/J.INSMATHECO.2025.103135MaRDI QIDQ6894252zbMATH OpenFDO

Authors Giles Hooker, Fei Huang

Publication date 25 November 2025

Published in Insurance Mathematics \& Economics (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/J.INSMATHECO.2025.103135

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Actuarial mathematics (91G05)

Cites work

• Boosting insights in insurance tariff plans with tree-based machine learning methods
• Causal Interpretations of Black-Box Models
• Delta Boosting Machine with Application to General Insurance
• Greedy function approximation: A gradient boosting machine.
• Insurance Premium Prediction via Gradient Tree-Boosted Tweedie Compound Poisson Models
• Prediction-based regularization using data augmented regression
• The Elements of Statistical Learning
• Unrestricted permutation forces extrapolation: variable importance requires at least one more model, or there is no free variable importance
• Visualizing the Effects of Predictor Variables in Black Box Supervised Learning Models

Cited in

(1)

• What KAN mortality say: smooth and interpretable mortality modeling using Kolmogorov-Arnold networks

This page was built for publication: Pitfalls in machine learning interpretability: manipulating partial dependence plots to hide discrimination

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6894252)