Local prediction on HAR (20% train)
1.9Local Prediction ErrorMean-reg
Evaluation results
| Method | Links | |
|---|---|---|
| Mean-regε (contamination level)=0.00, training data percentage=20%2026.07 | 1.9 | |
| robust multi-task gradient descentε (contamination level)=0.00, training data percentage=20%2026.07 | 2.1 | |
| robust multi-task gradient descentε (contamination level)=0.05, training data percentage=20%2026.07 | 2.1 | |
| robust multi-task gradient descentε (contamination level)=0.10, training data percentage=20%2026.07 | 2.2 | |
| robust multi-task gradient descentε (contamination level)=0.15, training data percentage=20%2026.07 | 2.2 | |
| robust multi-task gradient descentε (contamination level)=0.20, training data percentage=20%2026.07 | 2.3 | |
| robust multi-task gradient descentε (contamination level)=0.25, training data percentage=20%2026.07 | 2.3 | |
| ARMULε (contamination level)=0.00, training data percentage=20%2026.07 | 3.1 | |
| Mean-regε (contamination level)=0.05, training data percentage=20%2026.07 | 3.3 | |
| Averageε (contamination level)=0.00, training data percentage=20%2026.07 | 3.7 | |
| Trimmed meanε (contamination level)=0.00, training data percentage=20%2026.07 | 3.7 | |
| Bulyanε (contamination level)=0.00, training data percentage=20%2026.07 | 3.7 | |
| Filteringε (contamination level)=0.00, training data percentage=20%2026.07 | 3.7 | |
| MoM-Filteringε (contamination level)=0.00, training data percentage=20%2026.07 | 3.7 | |
| Historyε (contamination level)=0.00, training data percentage=20%2026.07 | 3.7 | |
| Bucketingε (contamination level)=0.00, training data percentage=20%2026.07 | 3.7 | |
| Filteringε (contamination level)=0.05, training data percentage=20%2026.07 | 3.8 | |
| MoM-Filteringε (contamination level)=0.05, training data percentage=20%2026.07 | 3.8 | |
| Trimmed meanε (contamination level)=0.05, training data percentage=20%2026.07 | 3.9 | |
| Medianε (contamination level)=0.00, training data percentage=20%2026.07 | 4.1 | |
| Trimmed meanε (contamination level)=0.10, training data percentage=20%2026.07 | 4.1 | |
| Historyε (contamination level)=0.05, training data percentage=20%2026.07 | 4.1 | |
| Bulyanε (contamination level)=0.05, training data percentage=20%2026.07 | 4.2 | |
| Medianε (contamination level)=0.05, training data percentage=20%2026.07 | 4.3 | |
| Trimmed meanε (contamination level)=0.15, training data percentage=20%2026.07 | 4.3 | |
| Filteringε (contamination level)=0.10, training data percentage=20%2026.07 | 4.3 | |
| MoM-Filteringε (contamination level)=0.10, training data percentage=20%2026.07 | 4.3 | |
| Medianε (contamination level)=0.10, training data percentage=20%2026.07 | 4.5 | |
| Medianε (contamination level)=0.15, training data percentage=20%2026.07 | 4.6 | |
| Trimmed meanε (contamination level)=0.20, training data percentage=20%2026.07 | 4.6 | |
| Dirtyε (contamination level)=0.00, training data percentage=20%2026.07 | 4.6 | |
| Trimmed meanε (contamination level)=0.25, training data percentage=20%2026.07 | 4.7 | |
| Filteringε (contamination level)=0.15, training data percentage=20%2026.07 | 4.7 | |
| MoM-Filteringε (contamination level)=0.15, training data percentage=20%2026.07 | 4.7 | |
| Averageε (contamination level)=0.05, training data percentage=20%2026.07 | 4.8 | |
| Medianε (contamination level)=0.20, training data percentage=20%2026.07 | 4.8 | |
| Historyε (contamination level)=0.10, training data percentage=20%2026.07 | 4.8 | |
| Bucketingε (contamination level)=0.05, training data percentage=20%2026.07 | 4.8 | |
| Medianε (contamination level)=0.25, training data percentage=20%2026.07 | 5 | |
| RMTFLε (contamination level)=0.00, training data percentage=20%2026.07 | 5.2 | |
| Bulyanε (contamination level)=0.10, training data percentage=20%2026.07 | 5.3 | |
| Historyε (contamination level)=0.15, training data percentage=20%2026.07 | 5.3 | |
| Dirtyε (contamination level)=0.10, training data percentage=20%2026.07 | 5.5 | |
| Dirtyε (contamination level)=0.20, training data percentage=20%2026.07 | 5.5 | |
| Dirtyε (contamination level)=0.25, training data percentage=20%2026.07 | 5.5 | |
| Filteringε (contamination level)=0.20, training data percentage=20%2026.07 | 5.6 | |
| MoM-Filteringε (contamination level)=0.20, training data percentage=20%2026.07 | 5.6 | |
| Dirtyε (contamination level)=0.05, training data percentage=20%2026.07 | 5.6 | |
| Dirtyε (contamination level)=0.15, training data percentage=20%2026.07 | 5.6 | |
| RMTFLε (contamination level)=0.05, training data percentage=20%2026.07 | 5.6 | |
| RMTFLε (contamination level)=0.10, training data percentage=20%2026.07 | 5.6 | |
| RMTFLε (contamination level)=0.15, training data percentage=20%2026.07 | 5.6 | |
| RMTFLε (contamination level)=0.20, training data percentage=20%2026.07 | 5.6 | |
| RMTFLε (contamination level)=0.25, training data percentage=20%2026.07 | 5.6 | |
| Bulyanε (contamination level)=0.15, training data percentage=20%2026.07 | 5.7 | |
| Mean-regε (contamination level)=0.10, training data percentage=20%2026.07 | 5.7 | |
| Single-taskε (contamination level)=0.00, training data percentage=20%2026.07 | 5.8 | |
| Single-taskε (contamination level)=0.05, training data percentage=20%2026.07 | 5.8 | |
| Single-taskε (contamination level)=0.10, training data percentage=20%2026.07 | 5.8 | |
| Single-taskε (contamination level)=0.15, training data percentage=20%2026.07 | 5.8 | |
| Single-taskε (contamination level)=0.20, training data percentage=20%2026.07 | 5.8 | |
| Mean-regε (contamination level)=0.15, training data percentage=20%2026.07 | 5.8 | |
| Mean-regε (contamination level)=0.20, training data percentage=20%2026.07 | 5.8 | |
| Mean-regε (contamination level)=0.25, training data percentage=20%2026.07 | 5.8 | |
| Single-taskε (contamination level)=0.25, training data percentage=20%2026.07 | 5.9 | |
| Bulyanε (contamination level)=0.20, training data percentage=20%2026.07 | 6.1 | |
| Bulyanε (contamination level)=0.25, training data percentage=20%2026.07 | 6.2 | |
| Filteringε (contamination level)=0.25, training data percentage=20%2026.07 | 6.2 | |
| MoM-Filteringε (contamination level)=0.25, training data percentage=20%2026.07 | 6.2 | |
| Historyε (contamination level)=0.20, training data percentage=20%2026.07 | 6.7 | |
| ARMULε (contamination level)=0.05, training data percentage=20%2026.07 | 7 | |
| Historyε (contamination level)=0.25, training data percentage=20%2026.07 | 7.7 | |
| Averageε (contamination level)=0.10, training data percentage=20%2026.07 | 7.9 | |
| Bucketingε (contamination level)=0.10, training data percentage=20%2026.07 | 7.9 | |
| Bucketingε (contamination level)=0.15, training data percentage=20%2026.07 | 8.1 | |
| Averageε (contamination level)=0.15, training data percentage=20%2026.07 | 9.9 | |
| Bucketingε (contamination level)=0.20, training data percentage=20%2026.07 | 11.5 | |
| Bucketingε (contamination level)=0.25, training data percentage=20%2026.07 | 13.2 | |
| Averageε (contamination level)=0.20, training data percentage=20%2026.07 | 13.4 | |
| Averageε (contamination level)=0.25, training data percentage=20%2026.07 | 14.9 | |
| ARMULε (contamination level)=0.10, training data percentage=20%2026.07 | 15.4 | |
| ARMULε (contamination level)=0.15, training data percentage=20%2026.07 | 16.5 | |
| ARMULε (contamination level)=0.20, training data percentage=20%2026.07 | 17.1 | |
| RLRMTLε (contamination level)=0.05, training data percentage=20%2026.07 | 17.1 | |
| RLRMTLε (contamination level)=0.10, training data percentage=20%2026.07 | 17.1 | |
| ARMULε (contamination level)=0.25, training data percentage=20%2026.07 | 17.2 | |
| RLRMTLε (contamination level)=0.00, training data percentage=20%2026.07 | 17.2 | |
| RLRMTLε (contamination level)=0.15, training data percentage=20%2026.07 | 17.2 | |
| RLRMTLε (contamination level)=0.20, training data percentage=20%2026.07 | 17.2 | |
| RLRMTLε (contamination level)=0.25, training data percentage=20%2026.07 | 17.2 | |
| MoM-Krumε (contamination level)=0.00, training data percentage=20%2026.07 | 19.1 | |
| MoM-Krumε (contamination level)=0.05, training data percentage=20%2026.07 | 19.1 | |
| MoM-Krumε (contamination level)=0.10, training data percentage=20%2026.07 | 19.3 | |
| MoM-Krumε (contamination level)=0.15, training data percentage=20%2026.07 | 19.3 | |
| MoM-Krumε (contamination level)=0.20, training data percentage=20%2026.07 | 20.2 | |
| MoM-Krumε (contamination level)=0.25, training data percentage=20%2026.07 | 20.6 | |
| Krumε (contamination level)=0.20, training data percentage=20%2026.07 | 30.3 | |
| Krumε (contamination level)=0.25, training data percentage=20%2026.07 | 30.3 | |
| Krumε (contamination level)=0.10, training data percentage=20%2026.07 | 30.5 | |
| Krumε (contamination level)=0.15, training data percentage=20%2026.07 | 30.5 |