Classification on Bank real-world (test)
90.6AccuracyRMB-CLE-via-LGBM
Evaluation Results
| Method | Links | |
|---|---|---|
| RMB-CLE-via-LGBMStrategy=Robust Multi-Task Boosting, Backbone=LightGBM2026.02 | 90.6 | |
| DP-LGBMStrategy=Data-Parallel, Backbone=LightGBM2026.02 | 89.8 | |
| TaF-LGBMStrategy=Task-at-a-Time, Backbone=LightGBM2026.02 | 89.8 | |
| ST-LGBMStrategy=Single-Task, Backbone=LightGBM2026.02 | 89.7 | |
| R-MTGBStrategy=Robust Multi-Task, Backbone=Gradient Boosting2026.02 | 89.5 | |
| RMB-CLE-via-MTGBStrategy=Robust Multi-Task Boosting, Backbone=Multi-Task Gradient Boosting2026.02 | 89.4 | |
| MTGBStrategy=Multi-Task, Backbone=Gradient Boosting2026.02 | 89.3 | |
| ST-GBStrategy=Single-Task, Backbone=Gradient Boosting2026.02 | 89.2 | |
| DP-GBStrategy=Data-Parallel, Backbone=Gradient Boosting2026.02 | 88.9 | |
| TaF-GBStrategy=Task-at-a-Time, Backbone=Gradient Boosting2026.02 | 88.9 |