{"title":"强盗反馈下单调博弈中的非耦合和收敛学习","authors":"Jing Dong, Baoxiang Wang, Yaoliang Yu","doi":"arxiv-2408.08395","DOIUrl":null,"url":null,"abstract":"We study the problem of no-regret learning algorithms for general monotone\nand smooth games and their last-iterate convergence properties. Specifically,\nwe investigate the problem under bandit feedback and strongly uncoupled\ndynamics, which allows modular development of the multi-player system that\napplies to a wide range of real applications. We propose a mirror-descent-based\nalgorithm, which converges in $O(T^{-1/4})$ and is also no-regret. The result\nis achieved by a dedicated use of two regularizations and the analysis of the\nfixed point thereof. The convergence rate is further improved to $O(T^{-1/2})$\nin the case of strongly monotone games. Motivated by practical tasks where the\ngame evolves over time, the algorithm is extended to time-varying monotone\ngames. We provide the first non-asymptotic result in converging monotone games\nand give improved results for equilibrium tracking games.","PeriodicalId":501316,"journal":{"name":"arXiv - CS - Computer Science and Game Theory","volume":"11 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-08-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Uncoupled and Convergent Learning in Monotone Games under Bandit Feedback\",\"authors\":\"Jing Dong, Baoxiang Wang, Yaoliang Yu\",\"doi\":\"arxiv-2408.08395\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"We study the problem of no-regret learning algorithms for general monotone\\nand smooth games and their last-iterate convergence properties. Specifically,\\nwe investigate the problem under bandit feedback and strongly uncoupled\\ndynamics, which allows modular development of the multi-player system that\\napplies to a wide range of real applications. We propose a mirror-descent-based\\nalgorithm, which converges in $O(T^{-1/4})$ and is also no-regret. The result\\nis achieved by a dedicated use of two regularizations and the analysis of the\\nfixed point thereof. The convergence rate is further improved to $O(T^{-1/2})$\\nin the case of strongly monotone games. Motivated by practical tasks where the\\ngame evolves over time, the algorithm is extended to time-varying monotone\\ngames. We provide the first non-asymptotic result in converging monotone games\\nand give improved results for equilibrium tracking games.\",\"PeriodicalId\":501316,\"journal\":{\"name\":\"arXiv - CS - Computer Science and Game Theory\",\"volume\":\"11 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-08-15\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Computer Science and Game Theory\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2408.08395\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Computer Science and Game Theory","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2408.08395","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Uncoupled and Convergent Learning in Monotone Games under Bandit Feedback
We study the problem of no-regret learning algorithms for general monotone
and smooth games and their last-iterate convergence properties. Specifically,
we investigate the problem under bandit feedback and strongly uncoupled
dynamics, which allows modular development of the multi-player system that
applies to a wide range of real applications. We propose a mirror-descent-based
algorithm, which converges in $O(T^{-1/4})$ and is also no-regret. The result
is achieved by a dedicated use of two regularizations and the analysis of the
fixed point thereof. The convergence rate is further improved to $O(T^{-1/2})$
in the case of strongly monotone games. Motivated by practical tasks where the
game evolves over time, the algorithm is extended to time-varying monotone
games. We provide the first non-asymptotic result in converging monotone games
and give improved results for equilibrium tracking games.