Background: Over the past decade, the potential advantages of employing deep learning models and leveraging auxiliary data in data-driven end-to-end (E2E) frameworks to enhance inventory decision-making have gained recognition. However, current approaches predominantly rely on feed-forward...