Expanding the Data Analysis Toolkit: Explainable AI, Causal Learning, and Time-Series Foundation Models – Daeil Jang

Recent advances in data science have expanded the scope of data analysis beyond prediction accuracy toward interpretability, causal understanding, and generalizable learning across complex data structures. This lecture introduces three emerging methodological approaches that can be directly leveraged in modern data analysis workflows.

First, the lecture presents explainable artificial intelligence (XAI) techniques, focusing on SHAP and its extension to time-series explainability, to illustrate how model predictions can be decomposed into meaningful variable- and time-specific contributions. Second, it introduces machine-learning and deep-learning–based causal inference models, highlighting how these methods move beyond association to estimate intervention effects and heterogeneous impacts while maintaining interpretability. Third, the lecture explores recent time-series foundation models—such as Lag-LLaMA and TabPFN-based approaches—that enable transferable learning across diverse time-series tasks with minimal task-specific training.

Rather than treating these approaches as isolated research trends, this lecture frames them as complementary analytical tools that address key questions in data analysis: What drives model predictions? What would change under intervention? And how can models generalize across time and settings? Through this integrated perspective, the lecture aims to provide practical insight into how these three methods can be applied to real-world data analysis and inspire new research and application opportunities.