Graph Machine Learning: An Overview
Graph Neural Networks (GNNs) are gaining attention in data science and machine learning but still remain poorly understood outside expert circles. To grasp this exciting approach, we must start with the broader field of Graph Machine Learning (GML). Many online resources talk about GNNs and GML as if they are interchangeable concepts or as if GNNs are a panacea approach that makes other GML approaches obsolete. This is simply not the case. One of GML’s primary purposes is to compress large sparse graph data structures to enable feasible prediction and inference. GNNs are one way to accomplish this, perhaps the most advanced way, but not the only way. Understanding this will help create a better foundation for future parts of this series, where we will cover specific types of GNNs and related GML approaches in more detail.
At its core, Graph machine learning (GML) is the application of machine learning to graphs specifically for predictive and prescriptive tasks. GML has a variety of use cases across supply chain, fraud detection, recommendations, customer 360, drug discovery, and more.
One of the best ways to understand GML is through the different types of ML tasks it can accomplish. I break this out for supervised and unsupervised learning below.
0 Comments