Single-loop Actor-Critic Optimization on Infinite-horizon discounted MDP
2Complexity Bound[42]
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| [42]Markovian Sampling=✗, Assumptions=Unif. mix. rates Exploration, Single Timescale=✗2026.05 | 2 | — | — | |
| [32]Markovian Sampling=✗, Assumptions=Exploration, Single Timescale=✓2026.05 | 2 | — | — | |
| Single-Loop Off-Policy Actor-CriticMarkovian Sampling=✓, Assumptions=Irreducibility, Single Timescale=✓2026.05 | 2 | — | — | |
| [68]Markovian Sampling=✓, Assumptions=Unif. mix. rates, Single Timescale=✗2026.05 | 2.5 | — | — | |
| [31]Markovian Sampling=✗, Assumptions=Exploration, Single Timescale=✓2026.05 | 3 | — | — | |
| [18]Markovian Sampling=✗, Assumptions=Unif. mix. rates, Single Timescale=✓2026.05 | 4 | — | — | |
| [28]Markovian Sampling=✓, Assumptions=Ergodicity, Single Timescale=✗2026.05 | 6 | — | — |