Bandit Regret Minimization on Wheel Bandit delta=0.95
8,211.9Mean Cumulative RegretGradient-Laplace
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gradient-Laplacek=5002026.05 | 8,211.9 | 355 | |
| Greedy-Laplacek=5002026.05 | 8,616.7 | 574.5 | |
| Greedy-Laplacek=1002026.05 | 21,568.7 | 5,895.4 | |
| NeuralLineark=N/A2026.05 | 30,266 | 6,572.7 | |
| Last kk=5002026.05 | 30,556.2 | 7,942.9 | |
| Subnet Diagonalk=5002026.05 | 33,796.4 | 7,464.6 | |
| Last kk=1002026.05 | 35,340.1 | 7,576.7 | |
| Gradient-Laplacek=1002026.05 | 38,177.4 | 6,904.8 | |
| MAP (no exploration)k=N/A2026.05 | 43,298.9 | 7,146.9 | |
| Subnet Diagonalk=1002026.05 | 47,184.9 | 7,275.3 |