A New Sample-Efficient PAC Reinforcement Learning Algorithm

被引：0

作者：

Zehfroosh, Ashkan ^{[1
]}

Tanner, Herbert G. ^{[1
]}

机构：

[1] Univ Delaware, Dept Mech Engn, Newark, DE 19716 USA

来源：

2020 28TH MEDITERRANEAN CONFERENCE ON CONTROL AND AUTOMATION (MED) | 2020年

关键词：

HUMAN-ROBOT INTERACTION; MDPS;

D O I：

10.1109/med48518.2020.9182985

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

This paper introduces a new hybrid PAC RL algorithm for MDPs, which intelligently maintains favorable features of its parents. The DDQ algorithm, integrates model-free and model-based learning approaches, preserving some advantages from both. A PAC analysis of the DDQ algorithm is presented and its sample complexity is explicitly bounded. Numerical results from a small-scale example motivated by work on human-robot interaction models corroborates the theoretical predictions on sample complexity.

引用

页码：788 / 793

页数：6

共 50 条

[1] Sample-Efficient Reinforcement Learning of Undercomplete POMDPs
Jin, Chi
Kakade, Sham M.
Krishnamurthy, Akshay
Liu, Qinghua
ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020, 2020, 33
[2] Sample-Efficient Proper PAC Learning with Approximate Differential Privacy
Ghazi, Badih
Golowich, Noah
Kumar, Ravi
Manurangsi, Pasin
STOC '21: PROCEEDINGS OF THE 53RD ANNUAL ACM SIGACT SYMPOSIUM ON THEORY OF COMPUTING, 2021, : 183 - 196
[3] Sample-Efficient Reinforcement Learning in the Presence of Exogenous Information
Efroni, Yonathan
Foster, Dylan J.
Misra, Dipendra
Krishnamurthy, Akshay
Langford, John
CONFERENCE ON LEARNING THEORY, VOL 178, 2022, 178
[4] Sample-efficient Reinforcement Learning in Robotic Table Tennis
Tebbe, Jonas
Krauch, Lukas
Gao, Yapeng
Zell, Andreas
2021 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2021), 2021, : 4171 - 4178
[5] Sample-efficient reinforcement learning for CERN accelerator control
Kain, Verena
Hirlander, Simon
Goddard, Brennan
Velotti, Francesco Maria
Porta, Giovanni Zevi Della
Bruchon, Niky
Valentino, Gianluca
PHYSICAL REVIEW ACCELERATORS AND BEAMS, 2020, 23 (12)
[6] Conditional Abstraction Trees for Sample-Efficient Reinforcement Learning
Dadvar, Mehdi
Nayyar, Rashmeet Kaur
Srivastava, Siddharth
UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, 2023, 216 : 485 - 495
[7] Sample-Efficient Reinforcement Learning for Pose Regulation of a Mobile Robot
Brescia, Walter
De Cicco, Luca
Mascolo, Saverio
2022 11TH INTERNATIONAL CONFERENCE ON CONTROL, AUTOMATION AND INFORMATION SCIENCES (ICCAIS), 2022, : 42 - 47
[8] Safe and Sample-Efficient Reinforcement Learning Algorithms for Factored Environments
Simao, Thiago D.
PROCEEDINGS OF THE TWENTY-EIGHTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE, 2019, : 6460 - 6461
[9] Sample-Efficient Deep Reinforcement Learning with Directed Associative Graph
Yang, Dujia
Qin, Xiaowei
Xu, Xiaodong
Li, Chensheng
Wei, Guo
CHINA COMMUNICATIONS, 2021, 18 (06) : 100 - 113
[10] Sample-Efficient Reinforcement Learning with Stochastic Ensemble Value Expansion
Buckman, Jacob
Hafner, Danijar
Tucker, George
Brevdo, Eugene
Lee, Honglak
ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 31 (NIPS 2018), 2018, 31

← 1 2 3 4 5 →