Multi-armed bandits with switching penalties

被引:35
|
作者
Asawa, M
Teneketzis, D
机构
[1] Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor
基金
美国国家科学基金会;
关键词
D O I
10.1109/9.486316
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The multi-armed bandit problem with switching penalties (switching cost and switching delays) is investigated. It is shown that under an optimal policy, decisions about the processor allocation need to be made only at stopping times that achieve an appropriate index, the well-known ''Gittins index'' or a ''switching index'' that is defined for switching cost and switching delays, An algorithm for the computation of the ''switching index'' is presented. Furthermore, sufficient conditions for optimality of allocation strategies, based on limited look-ahead techniques, are established, These conditions together with the above-mentioned feature of optimal scheduling policies simplify the search for an optimal allocation policy. For a special class of multi-armed bandits (scheduling of parallel queues with switching penalties and no arrivals), it is shown that the aforementioned property of optimal policies is sufficient to determine an optimal allocation strategy, In general, the determination of optimal allocation policies remains a difficult and challenging task.
引用
收藏
页码:328 / 348
页数:21
相关论文
共 50 条
  • [1] Multi-armed Bandits with Metric Switching Costs
    Guha, Sudipto
    Munagala, Kamesh
    AUTOMATA, LANGUAGES AND PROGRAMMING, PT II, PROCEEDINGS, 2009, 5556 : 496 - +
  • [2] On Kernelized Multi-armed Bandits
    Chowdhury, Sayak Ray
    Gopalan, Aditya
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 70, 2017, 70
  • [3] Multi-armed Bandits with Compensation
    Wang, Siwei
    Huang, Longbo
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 31 (NIPS 2018), 2018, 31
  • [4] Simple fixes that accommodate switching costs in multi-armed bandits
    Teymourian, Ehsan
    Yang, Jian
    EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2025, 320 (03) : 616 - 627
  • [5] Regional Multi-Armed Bandits
    Wang, Zhiyang
    Zhou, Ruida
    Shen, Cong
    INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 84, 2018, 84
  • [6] Federated Multi-Armed Bandits
    Shi, Chengshuai
    Shen, Cong
    THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2021, 35 : 9603 - 9611
  • [7] Multi-armed Bandits with Probing
    Elumar, Eray Can
    Tekin, Cem
    Yagan, Osman
    2024 IEEE INTERNATIONAL SYMPOSIUM ON INFORMATION THEORY, ISIT 2024, 2024, : 2080 - 2085
  • [8] Ballooning multi-armed bandits
    Ghalme, Ganesh
    Dhamal, Swapnil
    Jain, Shweta
    Gujar, Sujit
    Narahari, Y.
    ARTIFICIAL INTELLIGENCE, 2021, 296
  • [9] Self-Unaware Adversarial Multi-Armed Bandits With Switching Costs
    Alipour-Fanid, Amir
    Dabaghchian, Monireh
    Zeng, Kai
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2023, 34 (06) : 2908 - 2922
  • [10] Multi-armed bandits for performance marketing
    Gigli, Marco
    Stella, Fabio
    INTERNATIONAL JOURNAL OF DATA SCIENCE AND ANALYTICS, 2024,