Multiclass classification on Retrofit system Stage 2 multiclass inner join
70.2AccuracyAutoGluon
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| AutoGluonEnsemble configuration=Ens. L32026.06 | 70.2 | 65.4 | 34.8 | 93.4 | |
| LLM-STACKEDLLM utilization strategy=STACKED2026.06 | 68 | 62.6 | 28.1 | 88.8 | |
| Random ForestData sampling technique=SMOTE2026.06 | 61.1 | 62.1 | 39 | 91.7 | |
| XGBoostData sampling technique=SMOTE2026.06 | 60.1 | 61.4 | 38 | 91.4 | |
| CatBoostData sampling technique=SMOTE2026.06 | 58.9 | 60 | 35.6 | 90.6 | |
| LightGBMData sampling technique=SMOTE2026.06 | 58.2 | 60.5 | 38.7 | 91.7 | |
| LLM-EMBED LogRegLLM utilization strategy=EMBED, Classifier=Logistic Regression2026.06 | 48 | 52.1 | 32.7 | 86.9 | |
| Logistic RegressionData sampling technique=SMOTE2026.06 | 29.3 | 33.2 | 14.8 | 76.8 | |
| LLM-PROMPTEDLLM utilization strategy=PROMPTED, Backbone=Sonnet 42026.06 | 10 | 1.8 | 1.2 | 50 |