Random Search Neural Networks for Efficient and Expressive Graph Learning

TL;DR

This paper introduces random search neural networks (RSNNs) that guarantee full node coverage in graph sampling, improving expressivity and efficiency over traditional random walk neural networks (RWNNs), especially in sparse graphs.

Contribution

The paper proposes RSNNs that operate on random searches for complete node coverage, reducing sampling complexity and enhancing expressivity in graph learning tasks.

Findings

RSNNs require only O(log |V|) searches for full edge coverage in sparse graphs.

RSNNs are universal approximators when paired with sequence models.

Empirically, RSNNs outperform RWNNs on molecular and protein benchmarks, with up to 16× fewer samples.

Abstract

Random walk neural networks (RWNNs) have emerged as a promising approach for graph representation learning, leveraging recent advances in sequence models to process random walks. However, under realistic sampling constraints, RWNNs often fail to capture global structure even in small graphs due to incomplete node and edge coverage, limiting their expressivity. To address this, we propose \textit{random search neural networks} (RSNNs), which operate on random searches, each of which guarantees full node coverage. Theoretically, we demonstrate that in sparse graphs, only $O(\log |V|)$ searches are needed to achieve full edge coverage, substantially reducing sampling complexity compared to the $O(|V|)$ walks required by RWNNs (assuming walk lengths scale with graph size). Furthermore, when paired with universal sequence models, RSNNs are universal approximators. We lastly show RSNNs are probabilistically invariant to graph isomorphisms, ensuring their expectation is an isomorphism-invariant graph function. Empirically, RSNNs consistently outperform RWNNs on molecular and protein benchmarks, achieving comparable or superior performance with up to 16$\times$ fewer sampled sequences. Our work bridges theoretical and practical advances in random walk based approaches, offering an efficient and expressive framework for learning on sparse graphs.