Meta-Learning Based Few-Shot Graph-Level Anomaly Detection
Liting Li, Yumeng Wang, Yueheng Sun
TL;DR
The paper tackles graph-level anomaly detection under few-shot, label-scarce settings by introducing MA-GAD, which combines graph compression with a gradient-matching objective to preserve GNN performance on compressed graphs, a meta-learning initialization from auxiliary graphs to enable rapid adaptation, and a bias-aware anomaly loss to improve separability between normal and anomalous nodes. A MAML-inspired training loop and efficient optimizers (ANIL/Reptile) enable fast fine-tuning with minimal data. Empirical results on four real-world biochemical datasets show MA-GAD and its variants achieve superior performance for both graph-level and subgraph anomaly detection under few-shot conditions, with ablations confirming the critical role of meta-learning and graph condensation. The work offers a practical approach to robust, data-efficient graph anomaly detection in domains with scarce labels and noisy graphs, with implications for chemistry and related fields.
Abstract
Graph-level anomaly detection aims to identify anomalous graphs or subgraphs within graph datasets, playing a vital role in various fields such as fraud detection, review classification, and biochemistry. While Graph Neural Networks (GNNs) have made significant progress in this domain, existing methods rely heavily on large amounts of labeled data, which is often unavailable in real-world scenarios. Additionally, few-shot anomaly detection methods based on GNNs are prone to noise interference, resulting in poor embedding quality and reduced model robustness. To address these challenges, we propose a novel meta-learning-based graph-level anomaly detection framework (MA-GAD), incorporating a graph compression module that reduces the graph size, mitigating noise interference while retaining essential node information. We also leverage meta-learning to extract meta-anomaly information from similar networks, enabling the learning of an initialization model that can rapidly adapt to new tasks with limited samples. This improves the anomaly detection performance on target graphs, and a bias network is used to enhance the distinction between anomalous and normal nodes. Our experimental results, based on four real-world biochemical datasets, demonstrate that MA-GAD outperforms existing state-of-the-art methods in graph-level anomaly detection under few-shot conditions. Experiments on both graph anomaly and subgraph anomaly detection tasks validate the framework's effectiveness on real-world datasets.