Extending XReason: Formal Explanations for Adversarial Detection

2024-12-31

AI Summary

Existing model-agnostic explainability methods (e.g., SHAP, LIME) suffer from low fidelity, poor stability, and inability to detect adversarial misleading samples when interpreting LightGBM models. Method: This paper proposes XReason—a systematic extension introducing (i) the first formal explanation framework for LightGBM based on SAT solving, adapted from prior XGBoost formalizations; (ii) class-level formal explanations that yield verifiable semantic characterizations of model decision logic; and (iii) an explanation-driven co-generation and detection framework for adversarial samples. Results: Evaluated on CICIDS-2017, XReason achieves >98.5% attack detection accuracy, provides provably consistent explanations, and incurs only millisecond-scale inference overhead—significantly outperforming mainstream heuristic explainers in both fidelity and robustness.

Technology Category

• Humans and AI: Explainable AI (XAI) for Human Understanding
• Machine Learning: Adversarial Learning & Robustness
• Computer Vision: Adversarial Attacks & Robustness

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
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Explainable Artificial Intelligence (XAI) plays an important role in improving the transparency and reliability of complex machine learning models, especially in critical domains such as cybersecurity. Despite the prevalence of heuristic interpretation methods such as SHAP and LIME, these techniques often lack formal guarantees and may produce inconsistent local explanations. To fulfill this need, few tools have emerged that use formal methods to provide formal explanations. Among these, XReason uses a SAT solver to generate formal instance-level explanation for XGBoost models. In this paper, we extend the XReason tool to support LightGBM models as well as class-level explanations. Additionally, we implement a mechanism to generate and detect adversarial examples in XReason. We evaluate the efficiency and accuracy of our approach on the CICIDS-2017 dataset, a widely used benchmark for detecting network attacks.

Problem

• Interpretable AI
• Complex Machine Learning Models
• LightGBM Interpretation

Innovation

• Enhanced XReason Tool
• LightGBM Support
• Misleading Instance Detection