Abstract

Join us on Zoom

While Graph Neural Networks (GNNs) have recently become the de facto standard for modeling relational data, they impose a strong assumption on the availability of the node or edge features of the graph. In many real-world applications, however, features are only partially available; for example, in social networks, age and gender are available only for a small subset of users. We present a general approach for handling missing features in graph machine learning applications that is based on minimization of the Dirichlet energy and leads to a diffusion-type differential equation on the graph. The discretization of this equation produces a simple, fast and scalable algorithm which we call Feature Propagation. We experimentally show that the proposed approach outperforms previous methods on six common node-classification benchmarks and can withstand surprisingly high rates of missing features: on average we observe only around 4% relative accuracy drop when 99% of the features are missing. Moreover, it takes only 10 seconds to run on a graph with ∼2.5M nodes and ∼123M edges on a single GPU .

BIO : Emanuele is a Machine Learning Researcher at Twitter and a Ph.D. student at Imperial College London, working on Graph Neural Networks and supervised by Prof. Michael Bronstein. His research interests span a wide array of topics around graph neural networks, including scalability, dynamic graphs, and learning with missing node features. Before his current position, Emanuele was working at Fabula AI, which was then acquired by Twitter in June 2019. Previously, he completed an MPhil at the University of Cambridge and a BEng at Imperial College London, both in Computer Science.