An Automatic System Architecture Applying XAI for Dataset Feature Selection in Supervised Learning

Dataset feature selection is an important process in supervised learning, which directly affects the accuracy performance of the machine learning model. Traditional methods often prioritize accuracy over transparency, making it difficult to understand the importance of selected features. This paper proposes a new automated system architecture that applies Explainable AI (XAI) techniques to select dataset features in supervised learning. The proposed system architecture integrates XAI to ensure both high model accuracy and clear interpretability of selected features.