DeepAI AI Chat
Log In Sign Up

SHAP-IQ: Unified Approximation of any-order Shapley Interactions

by   Fabian Fumagalli, et al.

Predominately in explainable artificial intelligence (XAI) research, the Shapley value (SV) is applied to determine feature importance scores for any black box model. Shapley interaction indices extend the Shapley value to define any-order feature interaction scores. Defining a unique Shapley interaction index is an open research question and, so far, three definitions have been proposed, which differ by their choice of axioms. Moreover, each definition requires a specific approximation technique. We, however, propose SHAPley Interaction Quantification (SHAP-IQ), an efficient sampling-based approximator to compute Shapley interactions for all three definitions, as well as all other that satisfy the linearity, symmetry and dummy axiom. SHAP-IQ is based on a novel representation and, in contrast to existing methods, we provide theoretical guarantees for its approximation quality, as well as estimates for the variance of the point estimates. For the special case of SV, our approach reveals a novel representation of the SV and corresponds to Unbiased KernelSHAP with a greatly simplified calculation. We illustrate the computational efficiency and effectiveness by explaining state-of-the-art language models among high-dimensional synthetic models.


page 1

page 2

page 3

page 4


Faith-Shap: The Faithful Shapley Interaction Index

Shapley values, which were originally designed to assign attributions to...

Look at the Variance! Efficient Black-box Explanations with Sobol-based Sensitivity Analysis

We describe a novel attribution method which is grounded in Sensitivity ...

Non-asymptotic Results for Langevin Monte Carlo: Coordinate-wise and Black-box Sampling

Euler-Maruyama and Ozaki discretization of a continuous time diffusion p...

Exploring the cloud of feature interaction scores in a Rashomon set

Interactions among features are central to understanding the behavior of...

A New Interaction Index inspired by the Taylor Series

We study interactions among players in cooperative games. We propose a n...

Incremental Permutation Feature Importance (iPFI): Towards Online Explanations on Data Streams

Explainable Artificial Intelligence (XAI) has mainly focused on static l...