Online Learning (Dynamic Regret Minimization) on Adversarial Family Synthetic
20.64Mean Dynamic RegretPCGS
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PCGSHorizon (T)=600, Expert Pool Size (K)=32, Number of sequences (N)=202026.03 | 20.64 | 0.46 | |
| GenShare(heur)Horizon (T)=600, Expert Pool Size (K)=32, Number of sequences (N)=202026.03 | 21.19 | 0.42 | |
| FixedShareHorizon (T)=600, Expert Pool Size (K)=32, Number of sequences (N)=202026.03 | 21.39 | 0.41 | |
| HedgeHorizon (T)=600, Expert Pool Size (K)=32, Number of sequences (N)=202026.03 | 31.31 | 2.19 |