AI for Earth and Planetary Sciences

Harvard FAS

This graduate-level course provides a comprehensive overview of machine learning methods for applications in Earth and Planetary Science (EPS). Topics range from foundational techniques (e.g., regression, clustering, random forests) to advanced deep learning architectures (CNNs, RNNs, GNNs, Transformer, Gen AI, LLM, PINNs, FNO, Agentic AI, Symbolic Regression, XAI). The curriculum integrates methodological lectures with hands-on labs using real-world EPS datasets (e.g., satellite imagery,…