System Identification on Robot Arm
16.57RMSEDeep Ensemble
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Deep EnsembleBase Model=NODE, Coverage level=90%2026.05 | 16.57 | 94.09 | 43.74 | 0.44 | |
| Deep EnsembleBase Model=NODE, Coverage=95%2026.05 | 16.57 | 97.55 | 52.12 | 0.52 | |
| MC DropoutBase Model=NODE, Coverage level=90%2026.05 | 16.66 | 96.18 | 49.54 | 0.5 | |
| MC DropoutBase Model=NODE, Coverage=95%2026.05 | 16.66 | 98.63 | 59.03 | 0.59 | |
| J-INNBase Model=LSTM-2, Coverage level=90%2026.05 | 18.35 | 91.04 | 43.07 | 0.86 | |
| C-INNBase Model=NODE-2, Coverage level=90%2026.05 | 18.57 | 87.95 | 39.86 | 1.08 | |
| C-INNBase Model=NODE-2, Coverage=95%2026.05 | 18.57 | 94.81 | 48.23 | 0.78 | |
| MC DropoutBase Model=LSTM, Coverage level=90%2026.05 | 19.15 | 90.86 | 44.31 | 0.75 | |
| MC DropoutBase Model=LSTM, Coverage=95%2026.05 | 19.15 | 95.07 | 52.79 | 0.94 | |
| C-INNBase Model=LSTM-2, Coverage level=90%2026.05 | 19.39 | 89.75 | 42.31 | 1.28 | |
| C-INNBase Model=LSTM-2, Coverage=95%2026.05 | 19.39 | 95.62 | 52.23 | 0.79 | |
| Bayesian NNBase Model=NODE, Coverage level=90%2026.05 | 19.6 | 92.45 | 46.45 | 0.51 | |
| Bayesian NNBase Model=NODE, Coverage=95%2026.05 | 19.6 | 96.65 | 55.34 | 0.67 | |
| Bayesian NNBase Model=LSTM, Coverage level=90%2026.05 | 19.63 | 87.28 | 40.39 | 1.59 | |
| Bayesian NNBase Model=LSTM, Coverage=95%2026.05 | 19.63 | 92.72 | 48.12 | 1.54 | |
| J-INNBase Model=NODE-2, Coverage level=90%2026.05 | 19.82 | 89.2 | 43.04 | 0.91 | |
| J-INNBase Model=NODE-2, Coverage=95%2026.05 | 19.82 | 95.68 | 53.15 | 0.79 | |
| J-INNBase Model=LSTM-2, Coverage=95%2026.05 | 20.15 | 94.54 | 56.01 | 1.61 | |
| Deep EnsembleBase Model=LSTM, Coverage level=90%2026.05 | 20.81 | 96.03 | 55.13 | 0.55 | |
| Deep EnsembleBase Model=LSTM, Coverage=95%2026.05 | 20.81 | 98.67 | 65.69 | 0.66 |