BREAD: Baseline-Referenced Explanations for Anomaly Diagnosis

2026-08-11

AI Summary

This study addresses the lack of interpretable diagnostic methods in AI-driven anomaly detection that simultaneously achieve generality, accuracy, and scalability, particularly in high-dimensional nonlinear settings where performance is often degraded by noise. To overcome this limitation, the authors propose a model-agnostic, explainable approach grounded in a normal-baseline reference framework. By integrating statistical process monitoring with a baseline comparison strategy, the method delivers theoretically guaranteed feature attributions under mean-shift anomaly assumptions. Empirical evaluations across multiple synthetic and real-world case studies demonstrate that the proposed technique substantially outperforms existing approaches such as LIME, achieving higher diagnostic fidelity and accuracy while remaining compatible with a wide range of AI-based anomaly detection models.

Abstract

Artificial Intelligence (AI)-based prospective anomaly detection methods are increasingly deployed in high-dimensional and nonlinear settings. Among these approaches, AI-based statistical process monitoring (SPM) is widely used, providing a structured framework for prospective monitoring. Once an anomaly is detected, a diagnosis method is needed to identify the features driving the flagged observation away from normal behaviour. Traditional SPM diagnosis methods are typically designed for specific detection models and cannot be directly applied to AI-based methods. Model-agnostic explainable AI (XAI) offers a general framework for feature relevance explanation. However, existing methods suffer from scalability limitations or assign relevance to noise features, reducing diagnosis accuracy. We propose a scalable, baseline-referenced diagnosis method that uses both the anomalous observation and normal baseline information. We provide mathematical guarantees that under a mean-shift anomaly setting, the proposed method achieves higher faithfulness in detecting the features causing the anomaly compared to LIME. Simulation studies and a real-world case study validate the effectiveness of the proposed method and show that it generates more faithful and accurate diagnosis results for AI-based prospective anomaly detection methods.

Problem

• anomaly diagnosis
• explainable AI
• statistical process monitoring
• feature relevance
• AI-based anomaly detection

Innovation

• baseline-referenced explanation
• anomaly diagnosis
• explainable AI
• statistical process monitoring
• faithfulness