Credit Risk Prediction on Home Credit Default Risk
0.7816ROC AUCGNN-Enhanced LightGBM
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GNN-Enhanced LightGBMCategory=Hybrid Ensemble2026.01 | 0.7816 | 0.2807 | — | |
| LightGBM (Strong Tabular)Category=Tabular Baseline2026.01 | 0.769 | 0.254 | 4.06 | |
| STRIKECategory=Proposed Framework2026.04 | 0.7661 | — | — | |
| SR-1D-CNNCategory=Deep Learning / Neural Networks2026.04 | 0.7517 | — | — | |
| Relation-Aware Attentive Heterogeneous GNNCategory=Graph Neural Network2026.01 | 0.7506 | 0.2291 | 1.57 | |
| LightGBMCategory=Tree-Based Ensembles2026.04 | 0.7489 | — | — | |
| AdaBoostCategory=Tree-Based Ensembles2026.04 | 0.7486 | — | — | |
| GBDTCategory=Tree-Based Ensembles2026.04 | 0.7485 | — | — | |
| XGBoostCategory=Tree-Based Ensembles2026.04 | 0.7475 | — | — | |
| CatBoostCategory=Tree-Based Ensembles2026.04 | 0.7453 | — | — | |
| TabNetCategory=Deep Learning / Neural Networks2026.04 | 0.7437 | — | — | |
| DeepFMCategory=Deep Learning / Neural Networks2026.04 | 0.7431 | — | — | |
| DCN-V2Category=Deep Learning / Neural Networks2026.04 | 0.7429 | — | — | |
| 1D CNNCategory=Deep Learning / Neural Networks2026.04 | 0.741 | — | — | |
| Heterogeneous GraphSAGE (6-node hetero graph)Category=Graph Neural Network2026.01 | 0.74 | 0.2217 | 0.14 | |
| LRCategory=Traditional Baselines2026.04 | 0.7392 | — | — | |
| Logistic RegressionCategory=Tabular Baseline2026.01 | 0.739 | 0.216 | — | |
| RFCategory=Tree-Based Ensembles2026.04 | 0.7091 | — | — | |
| MLPCategory=Deep Learning / Neural Networks2026.04 | 0.6981 | — | — | |
| Contrastive Pretraining + Fine-tuning (GraphCL-style)Category=Graph Neural Network2026.01 | 0.6804 | 0.1618 | 7.93 | |
| DTCategory=Traditional Baselines2026.04 | 0.6073 | — | — |