Reward Estimation on Standard bus engine replacement simulation without dummy variables
0.12MAPEGLADIUS
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GLADIUSMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=50002025.02 | 0.12 | 0.06 | |
| GLADIUSMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=25002025.02 | 0.13 | 0.06 | |
| RustMethod Category=Oracle Baselines, No. of Trajectories (H=100)=50002025.02 | 0.4 | 0.06 | |
| ML-IRLMethod Category=Oracle Baselines, No. of Trajectories (H=100)=50002025.02 | 0.43 | 0.26 | |
| GLADIUSMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=10002025.02 | 0.52 | 0.22 | |
| GLADIUSMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=5002025.02 | 0.55 | 0.2 | |
| ML-IRLMethod Category=Oracle Baselines, No. of Trajectories (H=100)=25002025.02 | 0.62 | 0.35 | |
| ML-IRLMethod Category=Oracle Baselines, No. of Trajectories (H=100)=10002025.02 | 0.64 | 0.48 | |
| RustMethod Category=Oracle Baselines, No. of Trajectories (H=100)=25002025.02 | 0.68 | 0.22 | |
| RustMethod Category=Oracle Baselines, No. of Trajectories (H=100)=10002025.02 | 0.71 | 0.49 | |
| ML-IRLMethod Category=Oracle Baselines, No. of Trajectories (H=100)=5002025.02 | 0.84 | 0.59 | |
| GLADIUSMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=2502025.02 | 0.84 | 0.51 | |
| RustMethod Category=Oracle Baselines, No. of Trajectories (H=100)=5002025.02 | 0.9 | 0.56 | |
| ML-IRLMethod Category=Oracle Baselines, No. of Trajectories (H=100)=2502025.02 | 1.1 | 0.78 | |
| RustMethod Category=Oracle Baselines, No. of Trajectories (H=100)=2502025.02 | 1.37 | 0.77 | |
| SAmQMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=10002025.02 | 1.55 | 0.46 | |
| SAmQMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=5002025.02 | 3.13 | 0.86 | |
| GLADIUSMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=502025.02 | 3.44 | 1.28 | |
| RustMethod Category=Oracle Baselines, No. of Trajectories (H=100)=502025.02 | 3.62 | 1.7 | |
| ML-IRLMethod Category=Oracle Baselines, No. of Trajectories (H=100)=502025.02 | 3.62 | 1.74 | |
| SAmQMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=2502025.02 | 3.65 | 1 | |
| SAmQMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=502025.02 | 4.92 | 1.2 | |
| BCMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=50002025.02 | 46.45 | 8.22 | |
| BCMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=25002025.02 | 62.61 | 10.75 | |
| BCMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=5002025.02 | 71.92 | 12.44 | |
| BCMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=2502025.02 | 72.04 | 13.21 | |
| BCMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=10002025.02 | 72.17 | 12.11 | |
| BCMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=502025.02 | 80.55 | 12.82 | |
| IQ-learnMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=25002025.02 | 111.77 | 23.99 | |
| IQ-learnMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=2502025.02 | 112.86 | 27.31 | |
| IQ-learnMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=10002025.02 | 112.98 | 24.12 | |
| IQ-learnMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=5002025.02 | 113.27 | 25.54 | |
| IQ-learnMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=502025.02 | 114.13 | 26.6 | |
| IQ-learnMethod Category=Neural Network, No Knowledge of Transition Probabilities, No. of Trajectories (H=100)=50002025.02 | 119.18 | 22.55 |