{"type":"ScholarlyArticle","title":"Stability and Generalization for Decentralized Markov SGD","authors":[{"name":"Jiahuan Wang","slug":"jiahuan-wang","given":"Jiahuan","family":"Wang","affiliation":""},{"name":"Ziqing Wen","slug":"ziqing-wen","given":"Ziqing","family":"Wen","affiliation":""},{"name":"Ping Luo","slug":"ping-luo","given":"Ping","family":"Luo","affiliation":""},{"name":"Dongsheng Li","slug":"dongsheng-li","given":"Dongsheng","family":"Li","affiliation":""},{"name":"Tao Sun","slug":"tao-sun","given":"Tao","family":"Sun","affiliation":""}],"abstract":"Stochastic gradient methods are central to large-scale learning, yet their generalization theory typically relies on independent sampling assumptions. In many practical applications, data are generated by Markov chains and learning is performed in a decentralized manner, which introduces significant analytical challenges. In this work, we investigate the stability and generalization of decentralized stochastic gradient descent (SGD) and stochastic gradient descent ascent (SGDA) under Markov chain sampling. Leveraging a stability-based framework, we characterize how Markovian dependence and decentralized communication jointly influence generalization behavior. Our analysis captures the effects of network topology, Markov chain mixing properties, and primal–dual dynamics. We establish non-asymptotic generalization bounds for both algorithms, extending existing results on Markov stochastic gradient methods to decentralized and minimax settings.","doi":"10.24963/ijcai.2026/548","url":"https://ijcai.org/proceedings/2026/548","pdf":"https://ijcai.org/proceedings/2026/0548.pdf","markdown":"https://ijcai.org/proceedings/2026/548.md","bibtex":"https://ijcai.org/proceedings/2026/bibtex/548","volume":{"slug":"2026","title":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2026","url":"https://ijcai.org/proceedings/2026/","year":2026,"location":"Bremen, Germany","dates":"15-21 August 2026","editors":"Diego Calvanese","isbn":"978-1-956792-09-6","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2026"},"section":"Machine Learning","track":"Main Track","number":548,"pages":{"first":4922,"last":4930},"keywords":["Machine Learning: Learning theory","Machine Learning: Optimization"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}