Exponential Hardness of Reinforcement Learning with Linear Function Approximation

被引：0
作者：
Kane, Daniel [1 ]
Liu, Sihan [1 ]
Lovett, Shachar [1 ]
Mahajan, Gaurav [2 ]
Szepesvári, Csaba [3 ,4 ]
Weisz, Gellért [5 ]
机构：
[1] University of California, San Diego, United States
[2] Yale University, United States
[3] DeepMind, London, United Kingdom
[4] University of Alberta, Edmonton, Canada
[5] University College London, London, United Kingdom
来源：
Proceedings of Machine Learning Research | 2023年 / 195卷
关键词：
Compendex;
D O I：
36th Annual Conference on Learning Theory, COLT 2023
中图分类号：
学科分类号：
摘要：
Reinforcement learning
引用
收藏
页码：1588 / 1617
相关论文
共 50 条
  • [41] Approximation of the value function for optimal control problems on stratified domains
    Dipartimento di Matematica e Fisica, Università Roma Tre, Largo S. Leonardo Murialdo 1, Roma
    00146, Italy
    不详
    00161, Italy
    arXiv,
  • [42] USE OF FREQUENCY SPECIFICATION FUNCTIONS FOR TRANSFER FUNCTION APPROXIMATION.
    Shamash, Y.
    White, W.K.
    Journal Water Pollution Control Federation, 1980, : 588 - 591
  • [43] Optimal equidistant piecewise linear approximation algorithm for the computation of transcendental functions in FPGA
    Peng, Y. (pony911@163.com), 1600, Science Press (35):
  • [44] Research on Path Planning Strategy of Rescue Robot Based on Reinforcement Learning
    Shi, Ying-Ming
    Zhang, Zhiyuan
    Journal of Computers (Taiwan), 2022, 33 (03) : 187 - 194
  • [45] A Heuristically Accelerated Reinforcement Learning-Based Neurosurgical Path Planner
    Ji G.
    Gao Q.
    Zhang T.
    Cao L.
    Sun Z.
    Cyborg and Bionic Systems, 2023, 4
  • [46] Research on power allocation of integrated VLPC based on deep reinforcement learning
    Ma, Shuai
    Li, Bing
    Sheng, Haihong
    Gu, Rongyan
    Zhou, Hui
    Wang, Hongmei
    Wang, Yue
    Li, Shiyin
    Tongxin Xuebao/Journal on Communications, 2022, 43 (08): : 121 - 130
  • [47] Reinforcement learning-based tracking control for AUVs subject to disturbances
    Wang, Guangcang
    Zhang, Dianfeng
    Wu, Zhaojing
    Proceedings of the 34th Chinese Control and Decision Conference, CCDC 2022, 2022, : 2222 - 2227
  • [48] Training and Evaluation of Deep Policies Using Reinforcement Learning and Generative Models
    Ghadirzadeh, Ali
    Poklukar, Petra
    Arndt, Karol
    Finn, Chelsea
    Kyrki, Ville
    Kragic, Danica
    Björkman, Mårten
    Journal of Machine Learning Research, 2022, 23
  • [49] Cable SCARA Robot Controlled by a Neural Network Using Reinforcement Learning
    Okabe, Eduardo
    Paiva, Victor
    Silva-Teixeira, Luis H.
    Izuka, Jaime
    JOURNAL OF COMPUTATIONAL AND NONLINEAR DYNAMICS, 2023, 18 (10):
  • [50] Estimation on Human Motion Posture Using Improved Deep Reinforcement Learning
    Ma, Wenjing
    Zhao, Jianguang
    Zhu, Guangquan
    Journal of Computers (Taiwan), 2023, 34 (04) : 97 - 110
12345