Machine Learning Meets Panel Data: What Practitioners Need to Know

How to avoid overestimating machine learning models’ performance, usefulness, and real-world applicability due to hidden data leakage