FreeGAD: A Training-Free yet Effective Approach for Graph Anomaly Detection

TL;DR

FreeGAD introduces a training-free graph anomaly detection method that leverages affinity-gated residual encoding and anchor-guided scoring, achieving high performance without training costs.

Contribution

It proposes the first training-free GAD approach, reducing complexity and resource requirements while maintaining or improving detection accuracy.

Findings

Outperforms existing methods on multiple benchmarks.

Significantly reduces detection time and resource consumption.

Maintains high accuracy without training or iterative optimization.

Abstract

Graph Anomaly Detection (GAD) aims to identify nodes that deviate from the majority within a graph, playing a crucial role in applications such as social networks and e-commerce. Despite the current advancements in deep learning-based GAD, existing approaches often suffer from high deployment costs and poor scalability due to their complex and resource-intensive training processes. Surprisingly, our empirical findings suggest that the training phase of deep GAD methods, commonly perceived as crucial, may actually contribute less to anomaly detection performance than expected. Inspired by this, we propose FreeGAD, a novel training-free yet effective GAD method. Specifically, it leverages an affinity-gated residual encoder to generate anomaly-aware representations. Meanwhile, FreeGAD identifies anchor nodes as pseudo-normal and anomalous guides, followed by calculating anomaly scores through anchor-guided statistical deviations. Extensive experiments demonstrate that FreeGAD achieves superior anomaly detection performance, efficiency, and scalability on multiple benchmark datasets from diverse domains, without any training or iterative optimization.