Reinforcement Learning on Breakout Atari 2600 (test)
168Average Total RewardDQN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DQNpolicy=epsilon-greedy, epsilon=0.052013.12 | 168 | — | |
| Humanagent=expert human player2013.12 | 31 | — | |
| Contingencypolicy=epsilon-greedy, epsilon=0.052013.12 | 6 | — | |
| Sarsapolicy=epsilon-greedy, epsilon=0.052013.12 | 5.2 | — | |
| Randompolicy=random2013.12 | 1.2 | — | |
| DQN Bestpolicy=epsilon-greedy, epsilon=0.05, evaluation=best episode2013.12 | — | 225 | |
| HNeat Bestpolicy=deterministic2013.12 | — | 52 | |
| HNeat Pixelpolicy=deterministic, input=8 color channel representation2013.12 | — | 4 |