Lower Yield Strength Prediction on Steel Property Prediction
2.646MAE (%)TabPFN
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| TabPFNFine-tuning Protocol=m.a.f.t. (multitask adapter fine-tuning)2026.03 | 2.646 | 86.93 | 58.77 | 96.78 | 9.73 | |
| TabPFNFine-tuning Protocol=m.f.t. (multitask fine-tuning)2026.03 | 2.647 | 86.93 | 58.56 | 96.76 | 9.64 | |
| TabPFNFine-tuning Protocol=s.f.t. (single-task fine-tuning)2026.03 | 2.652 | 86.94 | 58.46 | 96.77 | 9.29 | |
| TabPFNFine-tuning Protocol=n.f.t. (no fine-tuning)2026.03 | 2.697 | 86.52 | 57.56 | 96.72 | 8.04 | |
| XGBoostAlgorithm=Gradient Boosting2026.03 | 2.732 | 85.96 | 57.89 | 96.53 | 6.68 | |
| TabPFNFine-tuning Protocol=m.a.f.t. (multitask adapter fine-tuning)2026.03 | 2.87 | 87.19 | 57.04 | 95.46 | 9.73 | |
| TabPFNFine-tuning Protocol=m.f.t. (multitask fine-tuning)2026.03 | 2.874 | 87.13 | 56.92 | 95.46 | 9.64 | |
| TabPFNFine-tuning Protocol=s.f.t. (single-task fine-tuning)2026.03 | 2.901 | 86.75 | 56.24 | 95.43 | 9.29 | |
| TabPFNFine-tuning Protocol=n.f.t. (no fine-tuning)2026.03 | 2.917 | 86.42 | 55.6 | 95.41 | 8.04 | |
| MultiTabLearning Strategy=Multi-Tabular Learning2026.03 | 2.922 | 83.89 | 53.74 | 96.26 | 0.21 | |
| PLELearning Strategy=Progressive Layered Extraction2026.03 | 2.924 | 84.26 | 53.82 | 96.06 | 0.61 | |
| STEMLearning Strategy=Shared-Task Enhanced Model2026.03 | 2.927 | 84.16 | 52.58 | 96.27 | -0.4 | |
| MMoELearning Strategy=Mixture-of-Experts2026.03 | 2.933 | 84.11 | 53.6 | 96.03 | 0.35 | |
| STLLearning Strategy=Single Task Learning2026.03 | 2.963 | 83.42 | 52.69 | 96.15 | 0 | |
| SAINTArchitecture=Transformer-based2026.03 | 2.964 | 83.15 | 52.45 | 96.25 | -0.77 | |
| XGBoostAlgorithm=Gradient Boosting2026.03 | 2.976 | 85.61 | 56.55 | 94.96 | 6.68 | |
| FT-TransformerArchitecture=Transformer-based2026.03 | 2.99 | 82.81 | 52.42 | 96.2 | -2.66 | |
| MTLLearning Strategy=Multi-Task Learning2026.03 | 3.026 | 83.09 | 50.16 | 96.11 | -2.91 | |
| MultiTabLearning Strategy=Multi-Tabular Learning2026.03 | 3.103 | 84.17 | 53 | 95.01 | 0.21 | |
| STEMLearning Strategy=Shared-Task Enhanced Model2026.03 | 3.126 | 83.87 | 52.19 | 94.67 | -0.4 | |
| MMoELearning Strategy=Mixture-of-Experts2026.03 | 3.148 | 84.08 | 52.81 | 94.89 | 0.35 | |
| PLELearning Strategy=Progressive Layered Extraction2026.03 | 3.164 | 83.52 | 52.96 | 94.57 | 0.61 | |
| STLLearning Strategy=Single Task Learning2026.03 | 3.18 | 84.58 | 51.58 | 94.85 | 0 | |
| MTLLearning Strategy=Multi-Task Learning2026.03 | 3.187 | 83.95 | 50.74 | 94.89 | -2.91 | |
| SAINTArchitecture=Transformer-based2026.03 | 3.207 | 82.75 | 51.72 | 94.81 | -0.77 | |
| FT-TransformerArchitecture=Transformer-based2026.03 | 3.22 | 82.51 | 51.49 | 94.82 | -2.66 |