Adaptively lowering the number of per-episode update epochs via a multi-armed bandit improves or matches PPO/DrAC returns on standard benchmarks while reducing training FLOPS.
For the model update phase, the computational overhead is Oupdate = Oforward + Obackward, (15) where Oforward = Obs1 ∗ B ∗ Nbatches ∗ Nupdate epochs (16) and Obackward = Oforward ∗
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Adaptive Data Exploitation in Deep Reinforcement Learning
Adaptively lowering the number of per-episode update epochs via a multi-armed bandit improves or matches PPO/DrAC returns on standard benchmarks while reducing training FLOPS.