Enhancing LIME using Neural Decision Trees

2026-03-21

AI Summary

This work addresses the limited fidelity of traditional LIME, which relies on surrogate models such as linear regression or decision trees that struggle to accurately capture the complex nonlinear local decision boundaries of black-box models. To overcome this limitation, the authors propose NDT-LIME, a novel approach that integrates neural decision trees (NDTs) into the LIME framework as the surrogate model. By leveraging the hierarchical structure and enhanced nonlinear representational capacity of NDTs, NDT-LIME achieves more accurate approximation of the black-box model’s local behavior while preserving interpretability. Experimental results across multiple standard tabular datasets demonstrate that NDT-LIME consistently and significantly outperforms conventional LIME in terms of explanation fidelity.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Reasoning under Uncertainty: Graphical Models

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Interpreting complex machine learning models is a critical challenge, especially for tabular data where model transparency is paramount. Local Interpretable Model-Agnostic Explanations (LIME) has been a very popular framework for interpretable machine learning, also inspiring many extensions. While traditional surrogate models used in LIME variants (e.g. linear regression and decision trees) offer a degree of stability, they can struggle to faithfully capture the complex non-linear decision boundaries that are inherent in many sophisticated black-box models. This work contributes toward bridging the gap between high predictive performance and interpretable decision-making. Specifically, we propose the NDT-LIME variant that integrates Neural Decision Trees (NDTs) as surrogate models. By leveraging the structured, hierarchical nature of NDTs, our approach aims at providing more accurate and meaningful local explanations. We evaluate its effectiveness on several benchmark tabular datasets, showing consistent improvements in explanation fidelity over traditional LIME surrogates.

Problem

• interpretability
• LIME
• non-linear decision boundaries
• surrogate models
• tabular data

Innovation

• Neural Decision Trees
• LIME
• Interpretable Machine Learning
• Surrogate Models
• Explanation Fidelity