{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/gossip-learning-with-linear-models-on-fully","title":"Gossip Learning with Linear Models on Fully Distributed Data","arxiv_id":"1109.1396","date":"2011-09-07","proceeding":null,"authors":["Róbert Ormándi","István Hegedüs","Márk Jelasity"],"abstract":"Machine learning over fully distributed data poses an important problem in\npeer-to-peer (P2P) applications. In this model we have one data record at each\nnetwork node, but without the possibility to move raw data due to privacy\nconsiderations. For example, user profiles, ratings, history, or sensor\nreadings can represent this case. This problem is difficult, because there is\nno possibility to learn local models, the system model offers almost no\nguarantees for reliability, yet the communication cost needs to be kept low.\nHere we propose gossip learning, a generic approach that is based on multiple\nmodels taking random walks over the network in parallel, while applying an\nonline learning algorithm to improve themselves, and getting combined via\nensemble learning methods. We present an instantiation of this approach for the\ncase of classification with linear models. Our main contribution is an ensemble\nlearning method which---through the continuous combination of the models in the\nnetwork---implements a virtual weighted voting mechanism over an exponential\nnumber of models at practically no extra cost as compared to independent random\nwalks. We prove the convergence of the method theoretically, and perform\nextensive experiments on benchmark datasets. Our experimental analysis\ndemonstrates the performance and robustness of the proposed approach.","url_abs":"http://arxiv.org/abs/1109.1396v3","url_pdf":"http://arxiv.org/pdf/1109.1396v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"gossip-learning-with-linear-models-on-fully","repo_url":"https://github.com/RobertOrmandi/Gossip-Learning-Framework","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"gossip-learning-with-linear-models-on-fully","repo_url":"https://github.com/ormandi/gossip-learning-framework","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"gossip-learning-with-linear-models-on-fully","repo_url":"https://github.com/rmadhwal/trustchain-superapp","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"gossip-learning-with-linear-models-on-fully","repo_url":"https://github.com/tribler/trustchain-superapp","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"ensemble-learning","task_name":"Ensemble Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1109.1396","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}