Shortest Path Finding on Shortest path synthetic degree-8 Gaussian linear
1.5Final Cumulative RegretDFHPG
Evaluation results
| Method | Links | |
|---|---|---|
| DFHPGFeedback regime=Pure bandit feedback, Horizon (T)=2,000, Replications=302026.05 | 1.5 | |
| DFHPG-0Feedback regime=Pure bandit feedback, Horizon (T)=2,000, Replications=302026.05 | 1.58 | |
| DFHPG-1Feedback regime=Pure bandit feedback, Horizon (T)=2,000, Replications=302026.05 | 1.81 | |
| GREEDYCBFeedback regime=Pure bandit feedback, Horizon (T)=2,000, Replications=302026.05 | 1.98 | |
| TSCBFeedback regime=Pure bandit feedback, Horizon (T)=2,000, Replications=302026.05 | 2 | |
| ϵ-GREEDYCBFeedback regime=Pure bandit feedback, Horizon (T)=2,000, Replications=302026.05 | 2.06 | |
| DFHPG-0Feedback regime=Semi-bandit feedback, Horizon (T)=2,000, Replications=302026.05 | 5.31 | |
| DFHPGFeedback regime=Semi-bandit feedback, Horizon (T)=2,000, Replications=302026.05 | 6.83 |