Boot Camp on Logic and Graph Learning

Explore the intersection of logic and graph learning in this comprehensive boot camp lecture delivered by Pablo Barcelo from Pontificia Universidad Catolica de Chile. Delve into the theoretical foundations that connect logical reasoning with graph-based machine learning approaches, examining how formal logic principles can enhance graph neural networks and related algorithms. Discover the mathematical frameworks that underpin both domains and learn how logical structures can be leveraged to improve graph learning tasks such as node classification, link prediction, and graph classification. Gain insights into the computational complexity aspects of combining logic with graph learning methods, and understand how theoretical computer science principles inform practical applications in this emerging field. The session provides a rigorous academic perspective on how logical inference mechanisms can be integrated with modern graph learning techniques to create more interpretable and robust machine learning models.