Temporal Graph Prototype-conditioned Conformal Prediction for Fraud Detection

2026-08-16

AI Summary

This study addresses the inefficiency of conformal prediction sets in temporal graph fraud detection caused by noise and class imbalance. To overcome these challenges, we propose ProtoCP, a novel framework that integrates prototype learning to suppress benign noise while incorporating neighborhood relative scoring and temporal diffusion mechanisms. These innovations effectively resolve calibration difficulties arising from class imbalance and data drift. Extensive experiments on four benchmark datasets demonstrate that ProtoCP consistently achieves target coverage rates while yielding significantly smaller prediction sets compared to state-of-the-art baselines. Consequently, this work substantially enhances both the calibration efficiency and stability of edge-level fraud detection in dynamic graph environments, offering a robust solution for reliable uncertainty quantification under complex temporal distributions.

Abstract

Conformal prediction (CP) provides distribution-free coverage guarantees and has emerged as a principled tool for uncertainty quantification. In edge-level fraud detection on temporal interaction graphs, where false positives and false negatives both carry substantial cost, such coverage guarantees are particularly appealing for risk-aware decision making. However, directly applying existing graph conformal predictors yields inefficient prediction sets due to two recurring properties of fraud data. Fraudulent interactions are often embedded in benign-dominated neighborhoods that dilute calibration signals, while extreme class imbalance leaves scarce labeled-fraud support in the calibration split and leads to overly conservative class-conditional thresholds. To address these issues, we propose ProtoCP, a conformal prediction framework for edge-level fraud detection on temporal graphs. ProtoCP improves calibration efficiency by focusing calibration on fraud-relevant subgraph context and producing more stable nonconformity scores under class imbalance and temporal drift. Specifically, it leverages learned prototypes to suppress benign-dominated noise in the calibration context and introduces a neighborhood-relative scoring mechanism with temporal score diffusion for stable class-conditional calibration. Experiments on four fraud benchmarks (YelpChi, S-FFSD, FTFD, and BankSim) show that ProtoCP achieves the target coverage with consistently smaller prediction sets than state-of-the-art baselines. Our codes are available at https://github.com/Picard1701ent/ProtoCP.git

Problem

• Conformal Prediction
• Fraud Detection
• Temporal Graphs
• Class Imbalance
• Prediction Set Efficiency

Innovation

• Conformal Prediction
• Temporal Graph
• Fraud Detection
• Graph Prototypes
• Temporal Score Diffusion