Short-duration arbitrage on ENTSO-E cross-zone 39 zones longest consecutive available
47.5Median Reward vs Optimal (%)absolute + relative + forecast DDQN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| absolute + relative + forecast DDQNPolicy=absolute + relative + forecast, seeds=52026.06 | 47.5 | 38.2 | |
| absolute + forecast DDQNPolicy=absolute + forecast, seeds=52026.06 | 40 | 32.4 | |
| absolute + relative DDQNPolicy=absolute + relative, seeds=52026.06 | 39.8 | 36 | |
| relative + forecast DDQNPolicy=relative + forecast, seeds=52026.06 | 33.3 | 23.8 | |
| Price-score heuristicPolicy=Price-score heuristic2026.06 | 29.2 | — | |
| forecast DDQNPolicy=forecast, seeds=52026.06 | 26.8 | 14 | |
| absolute DDQNPolicy=absolute, seeds=52026.06 | 5.7 | 1 | |
| relative DDQNPolicy=relative, seeds=52026.06 | 1.3 | 0.8 |