The Graph Lottery Ticket Hypothesis: Finding Sparse, Informative Graph Structure

Anton Tsitsulin, Bryan Perozzi

Keywords: graphs, graph neural networks, graph algorithms, sparsification

TL;DR: We can find extremely sparse "lottery tickets" in graphs–very sparse subsets of edges–that attain comparable or better performance to original graph structure.

Abstract:

Graph learning methods help utilize implicit relationships among data items, thereby reducing training label requirements and improving task performance. However, determining the optimal graph structure for a particular learning task remains a challenging research problem.

In this work, we introduce the Graph Lottery Ticket (GLT) Hypothesis – that there is an extremely sparse backbone for every graph, and that graph learning algorithms attain comparable performance when trained on that subgraph as on the full graph.
We identify and systematically study 8 key metrics of interest that directly influence the performance of graph learning algorithms.
Subsequently, we define the notion of a "winning ticket" for graph structure – an extremely sparse subset of edges that can deliver a robust approximation of the entire graph's performance. We propose a straightforward and efficient algorithm for finding these GLTs in arbitrary graphs. Empirically, we observe that performance of different graph learning algorithms can be matched or even exceeded on graphs with the average degree as low as 5.

Submission Number: 32