Three-way outcome classification on Fencing Dataset (Grouped Match-Pool Evaluation)
62.6AccuracyRF
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| RFInput Stream=2D pruned token summary, Feature Set=-blade summaries2025.09 | 62.6 | 62.8 | 63.4 | |
| Random forestInput Stream=2D pruned token summary, Feature Set=-interaction summary2025.09 | 62.4 | 63.2 | 65 | |
| RFInput Stream=2D pruned token summary, Feature Set=counts+move+hit2025.09 | 62.2 | 62.6 | 63.4 | |
| RFInput Stream=Lifted 3D ST-GCN token stream, Feature Set=pruned2025.09 | 62.2 | 55.7 | 41.2 | |
| RFInput Stream=2D pruned token summary, Feature Set=all families2025.09 | 61.4 | 61.4 | 61.5 | |
| HistGradientBoostingInput Stream=2D pruned token summary, Feature Set=pruned2025.09 | 59.9 | 56.1 | 47.1 | |
| Logistic regressionInput Stream=2D pruned token summary, Feature Set=pruned2025.09 | 54.1 | 51.2 | 44.1 |