Cumulative Regret Minimization on NIA Reward Function Spec A: saturation
333.5Median Cumulative RegretGibbs-TS
Evaluation Results
| Method | Links | |
|---|---|---|
| Gibbs-TSReward model specification=misspecified, T=2000, n=82026.05 | 333.5 | |
| ETC-TSReward model specification=misspecified, T=2000, n=82026.05 | 1,214.8 | |
| Agarwal et al.Reward model specification=well-specified, T=2000, n=82026.05 | 1,342.3 |