中文
相关论文

相关论文: Adversarial Fine-tuning in Offline-to-Online Reinf…

200 篇论文

We propose a novel adaptive reinforcement learning control approach for fault tolerant control of degrading systems that is not preceded by a fault detection and diagnosis step. Therefore, \textit{a priori} knowledge of faults that may…

系统与控制 · 电气工程与系统科学 2020-08-12 Ibrahim Ahmed , Marcos Quiñones-Grueiro , Gautam Biswas

We study offline meta-reinforcement learning, a practical reinforcement learning paradigm that learns from offline data to adapt to new tasks. The distribution of offline data is determined jointly by the behavior policy and the task.…

机器学习 · 计算机科学 2022-06-22 Haoqi Yuan , Zongqing Lu

Recently, robust reinforcement learning (RL) methods designed to handle adversarial input observations have received significant attention, motivated by RL's inherent vulnerabilities. While existing approaches have demonstrated reasonable…

机器学习 · 计算机科学 2025-06-23 Kosuke Nakanishi , Akihiro Kubo , Yuji Yasui , Shin Ishii

Off-policy learning plays a pivotal role in optimizing and evaluating policies prior to the online deployment. However, during the real-time serving, we observe varieties of interventions and constraints that cause inconsistency between the…

机器学习 · 计算机科学 2022-03-01 Da Xu , Yuting Ye , Chuanwei Ruan , Bo Yang

Reinforcement Learning (RL) has achieved remarkable success in sequential decision tasks. However, recent studies have revealed the vulnerability of RL policies to different perturbations, raising concerns about their effectiveness and…

机器学习 · 计算机科学 2025-07-08 Buqing Nie , Yangqing Fu , Jingtian Ji , Yue Gao

Learning control policies in simulation enables rapid, safe, and cost-effective development of advanced robotic capabilities. However, transferring these policies to the real world remains difficult due to the sim-to-real gap, where…

机器人学 · 计算机科学 2026-01-16 Jiahe Pan , Jiaxu Xing , Rudolf Reiter , Yifan Zhai , Elie Aljalbout , Davide Scaramuzza

Offline reinforcement learning enables agents to leverage large pre-collected datasets of environment transitions to learn control policies, circumventing the need for potentially expensive or unsafe online data collection. Significant…

机器学习 · 计算机科学 2022-03-17 Cong Lu , Philip J. Ball , Jack Parker-Holder , Michael A. Osborne , Stephen J. Roberts

We present an online model-based reinforcement learning algorithm suitable for controlling complex robotic systems directly in the real world. Unlike prevailing sim-to-real pipelines that rely on extensive offline simulation and model-free…

机器人学 · 计算机科学 2026-05-07 Fang Nan , Hao Ma , Qinghua Guan , Josie Hughes , Michael Muehlebach , Marco Hutter

One of the great promises of robot learning systems is that they will be able to learn from their mistakes and continuously adapt to ever-changing environments. Despite this potential, most of the robot learning systems today are deployed…

机器学习 · 计算机科学 2020-08-03 Ryan Julian , Benjamin Swanson , Gaurav S. Sukhatme , Sergey Levine , Chelsea Finn , Karol Hausman

The goal of offline reinforcement learning is to learn a policy from a fixed dataset, without further interactions with the environment. This setting will be an increasingly more important paradigm for real-world applications of…

机器人学 · 计算机科学 2020-11-17 Wenxuan Zhou , Sujay Bajracharya , David Held

Reinforcement learning (RL) is a powerful data-driven control method that has been largely explored in autonomous driving tasks. However, conventional RL approaches learn control policies through trial-and-error interactions with the…

机器人学 · 计算机科学 2021-11-03 Tianyu Shi , Dong Chen , Kaian Chen , Zhaojian Li

The performance of reinforcement learning depends upon designing an appropriate action space, where the effect of each action is measurable, yet, granular enough to permit flexible behavior. So far, this process involved non-trivial user…

机器学习 · 计算机科学 2021-06-08 Edoardo Cetin , Oya Celiktutan

Offline reinforcement learning (RL) allows learning sequential behavior from fixed datasets. Since offline datasets do not cover all possible situations, many methods collect additional data during online fine-tuning to improve performance.…

机器学习 · 计算机科学 2024-06-13 Mohammadreza Nakhaei , Aidan Scannell , Joni Pajarinen

We investigate robust model-free reinforcement learning algorithms designed for environments that may be dynamic or even adversarial. Traditional state-based policies often struggle to accommodate the challenges imposed by the presence of…

机器学习 · 计算机科学 2023-11-02 Udaya Ghai , Arushi Gupta , Wenhan Xia , Karan Singh , Elad Hazan

Recently, robust reinforcement learning (RL) methods against input observation have garnered significant attention and undergone rapid evolution due to RL's potential vulnerability. Although these advanced methods have achieved reasonable…

机器学习 · 计算机科学 2024-09-04 Kosuke Nakanishi , Akihiro Kubo , Yuji Yasui , Shin Ishii

We propose a novel model-based offline Reinforcement Learning (RL) framework, called Adversarial Model for Offline Reinforcement Learning (ARMOR), which can robustly learn policies to improve upon an arbitrary reference policy regardless of…

机器学习 · 计算机科学 2023-12-29 Mohak Bhardwaj , Tengyang Xie , Byron Boots , Nan Jiang , Ching-An Cheng

Offline reinforcement learning (RL) is crucial for real-world applications where exploration can be costly or unsafe. However, offline learned policies are often suboptimal, and further online fine-tuning is required. In this paper, we…

机器学习 · 计算机科学 2024-06-03 Hao Hu , Yiqin Yang , Jianing Ye , Chengjie Wu , Ziqing Mai , Yujing Hu , Tangjie Lv , Changjie Fan , Qianchuan Zhao , Chongjie Zhang

Model-based reinforcement learning has attracted much attention due to its high sample efficiency and is expected to be applied to real-world robotic applications. In the real world, as unobservable disturbances can lead to unexpected…

机器人学 · 计算机科学 2025-05-26 Taisuke Kobayashi

Learning policies from previously recorded data is a promising direction for real-world robotics tasks, as online learning is often infeasible. Dexterous manipulation in particular remains an open problem in its general form. The…

Robots have limited adaptation ability compared to humans and animals in the case of damage. However, robot damages are prevalent in real-world applications, especially for robots deployed in extreme environments. The fragility of robots…

机器人学 · 计算机科学 2020-12-01 Fan Yang , Chao Yang , Di Guo , Huaping Liu , Fuchun Sun