Decision trees constitute a simple yet powerful and interpretable machine learning tool. While tree-based methods are designed only for cross-sectional data, we propose an approach that combines decision trees with time series modeling and thereby bridges the gap between machine learning and...