Addressing Epistemic and Methodological Challenges in Empirical Machine Learning Research

Empirical research in machine learning faces significant epistemic and methodological challenges that undermine the reliability and replicability of research findings. A more comprehensive understanding of different types of empirical inquiry, including both exploratory and confirmatory approaches, is needed to improve the validity and impact of machine learning research.