Table 8 Performance comparison of HMOA-GNN and existing CCFD methods on IEEE-CIS Fraud Detection dataset.

From: HMOA-GNN: adaptive adversarial GraphSAGE with hierarchical hybrid sampling and metric-optimized graph construction for credit card fraud detection

Ref

Algorithm

Accuracy

Precision

Recall

F1-score

AUC

Zou et al.49

DAE

0.9030

0.2091

0.6970

0.3217

0.8957

Pang et al.50

DevNet

0.7398

0.0471

0.5976

0.0873

0.6739

Ileberi et al.51

GA-RF

0.9872

0.9863

0.5620

0.7160

0.9613

Ni et al.27

SOBT

0.9787

0.8819

0.4442

0.5908

0.7223

Xiang et al.52

GTAN

0.9700

0.7019

0.7172

0.7092

0.8762

Chang et al.53

TFD

0.9027

0.2318

0.8300

0.3624

0.9409

Ours

HMOA-GNN

0.9805

0.6627

0.8333

0.7383

0.9445

  1. Bold values indicate the best performance.
Back to article page