Abstract: Deep learning models are widely used in data-driven applications due to their high predictive performance, but their lack of interpretability limits their applicability in domains requiring ...