Z-Score Experience Replay in Off-Policy Deep Reinforcement Learning
Yana Yang, Meng Xi, Huiao Dai, Jiabao Wen, Jiachen Yang, Sergio Toral Marín · 2024
Read Z-Score Experience Replay in Off-Policy Deep Reinforcement Learning by Yana Yang, Meng Xi, Huiao Dai, Jiabao Wen, Jiachen Yang, Sergio Toral Marín (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11645091.txt.
Description
Yana Yang and co-authors improve how learning agents reuse their past experience, in a 2024 paper. Off-policy reinforcement learning stores interactions in a replay buffer and retrains on them later. Which memories get replayed, and how often, strongly shapes learning speed and final performance.
Their method scores each stored experience with z-score statistics and prioritizes replay accordingly. The shared surname team tested the scheme with a series of ablation experiments, removing components one at a time to check each one's contribution. The full method significantly improved the effectiveness of deep reinforcement learning algorithms, speeding convergence. The authors frame experience selection as a lever as important as the learning rule itself, and one that is cheap to adjust in existing systems.
Catalog record
| CREATOR | Yana Yang, Meng Xi, Huiao Dai, Jiabao Wen, Jiachen Yang, Sergio Toral Marín |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 50 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |