SWING: Unlocking Implicit Graph Representations for Graph Random Features
This work addresses a computational bottleneck for researchers and practitioners in machine learning who work with large-scale implicit graphs, offering an accelerator-friendly method that avoids graph materialization.
The paper tackles the problem of performing computations on graphs defined by implicit representations (i-graphs), such as ε-neighborhood graphs, by introducing SWING, a new algorithm that uses space walks and Gumbel-softmax sampling to approximate combinatorial calculations without materializing the graph, achieving efficient and accurate results as demonstrated in experiments.
We propose SWING: Space Walks for Implicit Network Graphs, a new class of algorithms for computations involving Graph Random Features on graphs given by implicit representations (i-graphs), where edge-weights are defined as bi-variate functions of feature vectors in the corresponding nodes. Those classes of graphs include several prominent examples, such as: $ε$-neighborhood graphs, used on regular basis in machine learning. Rather than conducting walks on graphs' nodes, those methods rely on walks in continuous spaces, in which those graphs are embedded. To accurately and efficiently approximate original combinatorial calculations, SWING applies customized Gumbel-softmax sampling mechanism with linearized kernels, obtained via random features coupled with importance sampling techniques. This algorithm is of its own interest. SWING relies on the deep connection between implicitly defined graphs and Fourier analysis, presented in this paper. SWING is accelerator-friendly and does not require input graph materialization. We provide detailed analysis of SWING and complement it with thorough experiments on different classes of i-graphs.