Off-Policy Evaluation on CIFAR-100
0.0007MSEMR
Evaluation Results
| Method | Links | |
|---|---|---|
| MRnumber of evaluation data (n)=1000, number of training data (m)=500, alpha* (target policy parameter)=0.62023.12 | 0.0007 | |
| DMnumber of evaluation data (n)=1000, number of training data (m)=500, alpha* (target policy parameter)=0.62023.12 | 0.022 | |
| DRosnumber of evaluation data (n)=1000, number of training data (m)=500, alpha* (target policy parameter)=0.62023.12 | 0.2628 | |
| SwitchDRnumber of evaluation data (n)=1000, number of training data (m)=500, alpha* (target policy parameter)=0.62023.12 | 1.1644 | |
| DRnumber of evaluation data (n)=1000, number of training data (m)=500, alpha* (target policy parameter)=0.62023.12 | 3,823.9 | |
| IPWnumber of evaluation data (n)=1000, number of training data (m)=500, alpha* (target policy parameter)=0.62023.12 | 4,116.9 |