{"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/good-semi-supervised-learning-that-requires-a","title":"Good Semi-supervised Learning that Requires a Bad GAN","arxiv_id":"1705.09783","date":"2017-05-27","proceeding":"NeurIPS 2017 12","authors":["Zihang Dai","Zhilin Yang","Fan Yang","William W. Cohen","Ruslan Salakhutdinov"],"abstract":"Semi-supervised learning methods based on generative adversarial networks\n(GANs) obtained strong empirical results, but it is not clear 1) how the\ndiscriminator benefits from joint training with a generator, and 2) why good\nsemi-supervised classification performance and a good generator cannot be\nobtained at the same time. Theoretically, we show that given the discriminator\nobjective, good semisupervised learning indeed requires a bad generator, and\npropose the definition of a preferred generator. Empirically, we derive a novel\nformulation based on our analysis that substantially improves over feature\nmatching GANs, obtaining state-of-the-art results on multiple benchmark\ndatasets.","url_abs":"http://arxiv.org/abs/1705.09783v3","url_pdf":"http://arxiv.org/pdf/1705.09783v3.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":"good-semi-supervised-learning-that-requires-a","repo_url":"https://github.com/kimiyoung/ssl_bad_gan","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"semi-supervised-image-classification","task_name":"Semi-Supervised Image Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/semi-supervised-image-classification-on-cifar","task":"Semi-Supervised Image Classification","dataset":"CIFAR-10, 4000 Labels","model":"Bad GAN","rank_in_archive_order":47,"of":49,"metrics":{"Percentage error":"14.41"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1705.09783","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}