图变换器欺诈检测：自监督预训练与保形风险控制
Graph-Transformer Fraud Detection with Self-Supervised Pretraining and Conformal Risk Control
- JP Morgan（摩根大通）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
针对企业交易网络中的协调欺诈，提出图变换器检测器GTFD，融合结构与时序证据，结合自监督预训练和保形风险控制，显著提升检测性能与鲁棒性。
AI中文摘要：
企业交易网络中的金融欺诈已变得更加协调且难以通过基于规则的引擎以及将每笔交易孤立对待的经典学习模型检测。本文提出了GTFD，一种图变换器欺诈检测器，融合了企业支付图中的结构性和时间性证据。GTFD使用多头图注意力网络编码图结构，使用门控变换器编码有序交易序列，并通过跨模态门控层结合两种视图。保形风险控制头将融合表示转换为具有有限样本覆盖保证的无阈值异常分数，网络通过自监督链接掩码预训练和对抗性增强进行训练，使其在标签稀缺和对抗性扰动下保持稳定。在一个包含协调欺诈团伙的企业交易基准上，GTFD达到了0.990的AUROC，F1分数为96.1%（精确率96.3%，召回率95.9%），准确率为98.4%。相对于最强基线，它将假阳性率降低了约29%，同时将协调欺诈团伙的召回率从85.1%提高到96.5%。消融实验将约2.0个AUROC点归因于自监督预训练，1.9个AUROC点归因于保形头，对抗性压力测试显示GTFD在扰动幅度0.20时保持89.2%的准确率，而次优模型降至76.4%。
英文摘要：
Financial fraud in corporate transaction networks has grown more coordinated and harder to detect with rule-based engines and with classical learning models that treat each transaction in isolation. This paper presents GTFD, a graph-transformer fraud detector that fuses structural and temporal evidence from a corporation's payment graph. GTFD encodes the graph with a multi-head graph attention network, encodes ordered transaction sequences with a gated transformer, and combines both views through a cross-modal gating layer. A conformal risk-control head converts the fused representation into threshold-free anomaly scores with finite-sample coverage guarantees, and the network is trained with self-supervised link-mask pretraining plus adversarial augmentation so it remains stable under scarce labels and under adversarial perturbation. On a corporate transaction benchmark enriched with coordinated fraud rings, GTFD reaches an AUROC of 0.990, an F1-score of 96.1% (precision 96.3%, recall 95.9%), and an accuracy of 98.4%. It reduces the false-positive rate by about 29% relative to the strongest baseline while raising coordinated fraud-ring recall from 85.1% to 96.5%. Ablations attribute roughly 2.0 AUROC points to self-supervised pretraining and 1.9 AUROC points to the conformal head, and adversarial stress tests show GTFD retains 89.2% accuracy at perturbation magnitude 0.20 where the next-best model falls to 76.4%.
