Priority Based Synchronization for Faster Learning in Games

被引:0
|
作者
Koochakzadeh, Abbasali [1 ]
Yazicioglu, Yasin [1 ]
机构
[1] Univ Minnesota, Dept Elect & Comp Engn, Minneapolis, MN 55455 USA
来源
2022 IEEE 61ST CONFERENCE ON DECISION AND CONTROL (CDC) | 2022年
关键词
D O I
10.1109/CDC51059.2022.9992513
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Learning in games has been widely used to solve many cooperative multi-agent problems such as coverage control, consensus, self-reconfiguration or vehicle-target assignment. One standard approach in this domain is to formulate the problem as a potential game and to use an algorithm such as log-linear learning to achieve the stochastic stability of globally optimal configurations. Standard versions of such learning algorithms are asynchronous, i.e., only one agent updates its action at each round of the learning process. To enable faster learning, we propose a synchronization strategy based on decentralized random prioritization of agents, which allows multiple agents to change their actions simultaneously when they do not affect each other's utility or feasible actions. We show that the proposed approach can be integrated into any standard asynchronous learning algorithm to improve the convergence speed while maintaining the limiting behavior (e.g., stochastically stable configurations). We support our theoretical results with simulations in a coverage control scenario.
引用
收藏
页码:2500 / 2505
页数:6
相关论文
共 50 条
  • [11] Equilibrium selection in signaling games with teams: Forward induction or faster adaptive learning?
    Cooper, David J.
    Kagel, John H.
    RESEARCH IN ECONOMICS, 2009, 63 (04) : 216 - 224
  • [12] Faster algorithms for finitary games
    Horn, Florian
    TOOLS AND ALGORITHMS FOR THE CONSTRUCTION AND ANALYSIS OF SYSTEMS, PROCEEDINGS, 2007, 4424 : 472 - +
  • [13] Priority Based Wireless Multi-Network Selection Games
    Hota, Mohit
    Kapoor, Sanjiv
    PROCEEDINGS OF THE 2018 THE NINETEENTH INTERNATIONAL SYMPOSIUM ON MOBILE AD HOC NETWORKING AND COMPUTING (MOBIHOC '18), 2018, : 290 - 299
  • [14] Priority and Synchronization Support for ROS
    Saito, Yukihiro
    Azumi, Takuya
    Kato, Shinpei
    Nishio, Nobuhiko
    PROCEEDINGS OF 2016 IEEE 4TH INTERNATIONAL CONFERENCE ON CYBER-PHYSICAL SYSTEMS, NETWORKS, AND APPLICATIONS (CPSNA), 2016, : 77 - 82
  • [15] Off-Policy Reinforcement Learning for Synchronization in Multiagent Graphical Games
    Li, Jinna
    Modares, Hamidreza
    Chai, Tianyou
    Lewis, Frank L.
    Xie, Lihua
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2017, 28 (10) : 2434 - 2445
  • [16] Priority and synchronization support for ROS
    1600, Institute of Electrical and Electronics Engineers Inc., United States
  • [17] Faster ILOD: Incremental learning for object detectors based on faster RCNN
    Peng, Can
    Zhao, Kun
    Lovell, Brian C.
    PATTERN RECOGNITION LETTERS, 2020, 140 : 109 - 115
  • [18] No-regret learning in games with noisy feedback: Faster rates and adaptivity via learning rate separation
    Hsieh, Yu-Guan
    Antonakopoulos, Kimon
    Cevher, Volkan
    Mertikopoulos, Panayotis
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35, NEURIPS 2022, 2022,
  • [19] Faster Algorithm for Turn-based Stochastic Games with Bounded Treewidth
    Chatterjee, Krishnendu
    Meggendorfer, Tobias
    Saona, Raimundo
    Svoboda, Jakub
    PROCEEDINGS OF THE 2023 ANNUAL ACM-SIAM SYMPOSIUM ON DISCRETE ALGORITHMS, SODA, 2023, : 4590 - 4605
  • [20] New axiomatizations of the Priority value for cooperative games with priority structure
    He, Songtao
    Shan, Erfang
    Sun, Yuxin
    OPERATIONS RESEARCH LETTERS, 2025, 60