LTL-guided Reinforcement Learning on Zones Infinite-horizon (test)
914µaccDEEPLTL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DEEPLTLTask Specification Type=REACH-STAY[5], |Q|=34, |δ|=1162026.02 | 914 | — | |
| SEMLTLTask Specification Type=SMALL, |Q|=4, |δ|=102026.02 | 18.65 | 3.59 | |
| DEEPLTLTask Specification Type=SMALL, |Q|=4, |δ|=102026.02 | 14.5 | — | |
| SEMLTLTask Specification Type=ALWAYS-REACTIVE[5, 1], |Q|=38, |δ|=2042026.02 | 3.73 | 7 | |
| SEMLTLTask Specification Type=COMPLEX-PATROL[5, 5], |Q|=45, |δ|=2012026.02 | 2.44 | 6.33 | |
| DEEPLTLTask Specification Type=COMPLEX-PATROL[5, 5], |Q|=45, |δ|=2012026.02 | 2.13 | — | |
| SEMLTLTask Specification Type=REACH-STAY[5], |Q|=34, |δ|=1162026.02 | 2 | 7.29 |