Correlation-Aware Feature Attribution Based Explainable AI

2025-11-20

AI Summary

Existing global attribution methods suffer from high computational overhead, poor stability under input-dependent perturbations, and limited scalability to large-scale heterogeneous data. To address these challenges, we propose two novel attribution frameworks: ExCIR and BlockCIR. ExCIR introduces a correlation-aware mechanism to mitigate redundant scoring, while BlockCIR robustly centers features and outputs to quantify sign-aligned co-variation, enabling block-level attribution via either predefined or data-driven feature grouping. Together, they support lightweight, transferable explanations on subpopulations. Empirically, both methods achieve top-k ranking stability comparable to state-of-the-art baselines across multimodal datasets, while reducing computational cost significantly. Their design ensures practical scalability and deployability in real-world applications involving diverse, high-dimensional data.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Machine Learning: Calibration & Uncertainty Quantification
• Humans and AI: Explainable AI (XAI) for Human Understanding

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Explainable AI (XAI) is increasingly essential as modern models become more complex and high-stakes applications demand transparency, trust, and regulatory compliance. Existing global attribution methods often incur high computational costs, lack stability under correlated inputs, and fail to scale efficiently to large or heterogeneous datasets. We address these gaps with emph{ExCIR} (Explainability through Correlation Impact Ratio), a correlation-aware attribution score equipped with a lightweight transfer protocol that reproduces full-model rankings using only a fraction of the data. ExCIR quantifies sign-aligned co-movement between features and model outputs after emph{robust centering} (subtracting a robust location estimate, e.g., median or mid-mean, from features and outputs). We further introduce extsc{BlockCIR}, a emph{groupwise} extension of ExCIR that scores emph{sets} of correlated features as a single unit. By aggregating the same signed-co-movement numerators and magnitudes over predefined or data-driven groups, extsc{BlockCIR} mitigates double-counting in collinear clusters (e.g., synonyms or duplicated sensors) and yields smoother, more stable rankings when strong dependencies are present. Across diverse text, tabular, signal, and image datasets, ExCIR shows trustworthy agreement with established global baselines and the full model, delivers consistent top-$k$ rankings across settings, and reduces runtime via lightweight evaluation on a subset of rows. Overall, ExCIR provides emph{computationally efficient}, emph{consistent}, and emph{scalable} explainability for real-world deployment.

Problem

• Addresses computational inefficiency in global feature attribution methods
• Improves stability of explanations under correlated input features
• Enables scalable explainability for large and heterogeneous datasets

Innovation

• Correlation-aware attribution score with lightweight transfer protocol
• Robust centering for quantifying feature-output co-movement
• Groupwise extension for scoring correlated feature sets