Ultimate Tensile Strength Prediction on Steel Property Prediction
1.701MAE (%)TabPFN
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| TabPFNFine-tuning Protocol=s.f.t. (single-task fine-tuning)2026.03 | 1.701 | 96.26 | 78.14 | 97.83 | 9.29 | |
| TabPFNFine-tuning Protocol=m.f.t. (multitask fine-tuning)2026.03 | 1.704 | 96.17 | 77.79 | 97.82 | 9.64 | |
| TabPFNFine-tuning Protocol=m.a.f.t. (multitask adapter fine-tuning)2026.03 | 1.705 | 96.23 | 77.92 | 97.82 | 9.73 | |
| TabPFNFine-tuning Protocol=n.f.t. (no fine-tuning)2026.03 | 1.751 | 95.94 | 76.75 | 97.75 | 8.04 | |
| XGBoostAlgorithm=Gradient Boosting2026.03 | 1.78 | 95.59 | 76.42 | 97.64 | 6.68 | |
| PLELearning Strategy=Progressive Layered Extraction2026.03 | 1.98 | 94.59 | 71.2 | 97.15 | 0.61 | |
| MultiTabLearning Strategy=Multi-Tabular Learning2026.03 | 1.992 | 94.86 | 70.85 | 97.25 | 0.21 | |
| STLLearning Strategy=Single Task Learning2026.03 | 2.012 | 94.81 | 70.16 | 96.84 | 0 | |
| STEMLearning Strategy=Shared-Task Enhanced Model2026.03 | 2.014 | 94.19 | 69.95 | 97.15 | -0.4 | |
| MMoELearning Strategy=Mixture-of-Experts2026.03 | 2.018 | 94.32 | 70.46 | 97.15 | 0.35 | |
| SAINTArchitecture=Transformer-based2026.03 | 2.021 | 94.57 | 70.03 | 97.18 | -0.77 | |
| MTLLearning Strategy=Multi-Task Learning2026.03 | 2.099 | 93.79 | 67.88 | 97.05 | -2.91 | |
| FT-TransformerArchitecture=Transformer-based2026.03 | 2.159 | 92.91 | 66.39 | 97.02 | -2.66 |