{"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/generalization-in-machine-learning-via","title":"Generalization in Machine Learning via Analytical Learning Theory","arxiv_id":"1802.07426","date":"2018-02-21","proceeding":null,"authors":["Kenji Kawaguchi","Yoshua Bengio","Vikas Verma","Leslie Pack Kaelbling"],"abstract":"This paper introduces a novel measure-theoretic theory for machine learning\nthat does not require statistical assumptions. Based on this theory, a new\nregularization method in deep learning is derived and shown to outperform\nprevious methods in CIFAR-10, CIFAR-100, and SVHN. Moreover, the proposed\ntheory provides a theoretical basis for a family of practically successful\nregularization methods in deep learning. We discuss several consequences of our\nresults on one-shot learning, representation learning, deep learning, and\ncurriculum learning. Unlike statistical learning theory, the proposed learning\ntheory analyzes each problem instance individually via measure theory, rather\nthan a set of problem instances via statistics. As a result, it provides\ndifferent types of results and insights when compared to statistical learning\ntheory.","url_abs":"http://arxiv.org/abs/1802.07426v3","url_pdf":"http://arxiv.org/pdf/1802.07426v3.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":"generalization-in-machine-learning-via","repo_url":"https://github.com/Learning-and-Intelligent-Systems/Analytical-Learning-Theory","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"generalization-in-machine-learning-via","repo_url":"https://github.com/Learning-and-Intelligent-Systems/DualCutout","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"machine-learning","task_name":"BIG-bench Machine Learning"},{"task_slug":"deep-learning","task_name":"Deep Learning"},{"task_slug":"learning-theory","task_name":"Learning Theory"},{"task_slug":"one-shot-learning","task_name":"One-Shot Learning"},{"task_slug":"representation-learning","task_name":"Representation Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1802.07426","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}