Unified method improves explainability of ai models in healthcare and finance
A unified framework for global and local interpretability using adaptive derivative-ordered random explanation
Machine LearningArtificial Intelligence
Summary
Understanding how complex machine learning models make decisions is very important, especially in areas like healthcare and finance. The authors introduce ADORE, a new explanation method that uses math techniques to better understand and explain these models. It combines overall feature importance with explanations for individual cases, capturing how features interact in complex ways. ADORE is also faster and works across different types of data like tables, text, and images. The authors provide an open-source tool so others can use their method easily.
What this means in practice
- •For healthcare data teams: Use ADORE to identify which patient features most influence complex diagnostic models and understand individual patient predictions for better clinical decisions.
- •For financial risk analysts: Apply ADORE to explain risk models by linking global feature effects with local transactions, helping to detect critical factors driving decisions.
Authors
Lemen Chao, Ming Lei, Anran Fanga
Abstract
The interpretability of complex machine learning models is of paramount importance, especially in real-world high-stakes domains such as healthcare and finance. However, existing post-hoc interpretability methods suffer from inherent limitations: fragmented analytical processes, inadequate capacity to model nonlinear feature interactions, computational inefficiencies, and over-reliance on specific model architectures. To address these challenges, this paper provides a novel method - Adaptive Derivative-Ordered Random Explanation (ADORE) - that leverages first- and second-order derivatives to accommodate nonlinear model complexities, while enabling effective capture of feature-sample interactions within a unified analytical framework. ADORE integrates global feature importance with local sample contributions, precisely quantifying feature impact by capturing both magnitude and direction, and identifying critical samples influencing model decisions. Furthermore, it achieves computational efficiency through randomized singular value decomposition (SVD) and dynamic sparsity detection, making it scalable to large, high-dimensional datasets. Experiments across three data modalities - tabular, text, and image - demonstrate that ADORE outperforms existing methods such as LIME and SHAP in handling complex interactions and computational efficiency, while providing detailed and reliable explanations. To facilitate adoption and reproducibility, ADORE has been released as an open-source Python package, hosted on GitHub, enabling researchers and practitioners to readily adapt and apply our approach to their specific tasks, models, and datasets.