Practical Geometric Deep Learning in Python

Explore the rapidly evolving field of Geometric Deep Learning (GDL) in this informative conference talk from YOW! 2019. Discover how GDL leverages network structures in data to enhance machine learning outcomes across various domains, including social science, medicine, and finance. Learn about the Graph Convolution Operator, the core component driving GDL, and its role in building deep learning models for network-structured data. Gain insights into StellarGraph, an open-source Python library developed by CSIRO's Data61, designed to make GDL algorithms easily accessible to data scientists and machine learning practitioners. Delve into StellarGraph's design philosophy, API, and analytics workflow, and see practical demonstrations of its application in product recommendation and social network moderation. Explore the challenges of designing and implementing a library for this rapidly evolving field of machine learning, and gain valuable knowledge to enhance your understanding of Geometric Deep Learning and its real-world applications.

Practical Geometric Deep Learning in Python • Pantelis Elinas • YOW! 2019