Graph Machine Learning: From Classical Approach to Graph Neural Network

Over the last few years, we've seen a rise in graph algorithms in a lot of use cases. One overlooked problem is that we lack a map to orient ourselves in this changing technological world. In this talk, we'll explain the logical steps and algorithms used for graph-based machine learning paths. You'll go on a journey starting with classical machine learning with hand-written graph features and machine learning models, moving on to node embedding starting from node2vec and arriving at graphSage, and finally reaching graph neural networks.


  • Julien Genovese, Graph Data Scientist at Data Reply