バランスリプレイと悲観的Qアンサンブルによるオフラインからオンラインへの強化学習
Offline-to-Online Reinforcement Learning via Balanced Replay and Pessimistic Q-Ensemble
オフライン強化学習で得た方策をオンライン微調整する際の分布シフトによる誤差を、オンライン優先とオフライン近傍サンプルを併用するバランスリプレイと悲観的Qアンサンブルで抑え、歩行・操作タスクでの効率と性能を向上させた論文。
著者: Seunghyun Lee, Younggyo Seo, Kimin Lee, Pieter Abbeel, Jinwoo Shin
分類: cs.RO, cs.LG
原文アブストラクト
Recent advance in deep offline reinforcement learning (RL) has made it possible to train strong robotic agents from offline datasets. However, depending on the quality of the trained agents and the application being considered, it is often desirable to fine-tune such agents via further online interactions. In this paper, we observe that state-action distribution shift may lead to severe bootstrap error during fine-tuning, which destroys the good initial policy obtained via offline RL. To address this issue, we first propose a balanced replay scheme that prioritizes samples encountered online while also encouraging the use of near-on-policy samples from the offline dataset. Furthermore, we leverage multiple Q-functions trained pessimistically offline, thereby preventing overoptimism concerning unfamiliar actions at novel states during the initial training phase. We show that the proposed method improves sample-efficiency and final performance of the fine-tuned robotic agents on various locomotion and manipulation tasks. Our code is available at: https://github.com/shlee94/Off2OnRL.
