XAI for CPS - Code and Results

These artefacts are utilised within the publication "Explainable AI to Improve Machine Learning Reliability for Industrial Cyber-Physical Systems". This article explores the use of explainable AI (XAI) techniques to evaluate machine learning models for a CPS use case. Specifically the C-SHAP method is applied to interpret CNN predictions on time series data. The use case involves detecting anomalies (Normal, NoFan, UnderVolt conditions) on an ARM-based computing platform. The data used by the code can be found here: https://doi.org/10.5281/zenodo.19135488. Refer to the provided README file for further details.