![[2601.18608] PolySHAP: Extending KernelSHAP with Interaction-Informed Polynomial Regression](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2601.18608] PolySHAP: Extending KernelSHAP with Interaction-Informed Polynomial Regression
Summary
The paper introduces PolySHAP, an extension of KernelSHAP that uses interaction-informed polynomial regression to improve the accuracy of Shapley value estimates in explainable AI.
Why It Matters
As Shapley values are crucial for understanding model predictions in AI, PolySHAP's enhancements can lead to more accurate interpretations of complex models, thereby improving trust and transparency in AI systems. This advancement is particularly relevant for practitioners in AI and machine learning who rely on explainability.
Key Takeaways
- PolySHAP improves upon KernelSHAP by using higher degree polynomials for better Shapley value estimation.
- The method captures non-linear interactions between features, enhancing the interpretability of AI models.
- Empirical results show that PolySHAP provides more accurate estimates across various benchmark datasets.
- The paper connects PolySHAP to paired sampling, offering theoretical justification for its practical performance.
- This advancement supports the growing need for explainable AI in critical applications.
Computer Science > Artificial Intelligence arXiv:2601.18608 (cs) [Submitted on 26 Jan 2026 (v1), last revised 17 Feb 2026 (this version, v2)] Title:PolySHAP: Extending KernelSHAP with Interaction-Informed Polynomial Regression Authors:Fabian Fumagalli, R. Teal Witter, Christopher Musco View a PDF of the paper titled PolySHAP: Extending KernelSHAP with Interaction-Informed Polynomial Regression, by Fabian Fumagalli and 2 other authors View PDF HTML (experimental) Abstract:Shapley values have emerged as a central game-theoretic tool in explainable AI (XAI). However, computing Shapley values exactly requires $2^d$ game evaluations for a model with $d$ features. Lundberg and Lee's KernelSHAP algorithm has emerged as a leading method for avoiding this exponential cost. KernelSHAP approximates Shapley values by approximating the game as a linear function, which is fit using a small number of game evaluations for random feature subsets. In this work, we extend KernelSHAP by approximating the game via higher degree polynomials, which capture non-linear interactions between features. Our resulting PolySHAP method yields empirically better Shapley value estimates for various benchmark datasets, and we prove that these estimates are consistent. Moreover, we connect our approach to paired sampling (antithetic sampling), a ubiquitous modification to KernelSHAP that improves empirical accuracy. We prove that paired sampling outputs exactly the same Shapley value approximations as second...
