{"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/virtual-adversarial-ladder-networks-for-semi","title":"Virtual Adversarial Ladder Networks For Semi-supervised Learning","arxiv_id":"1711.07476","date":"2017-11-20","proceeding":null,"authors":["Saki Shinoda","Daniel E. Worrall","Gabriel J. Brostow"],"abstract":"Semi-supervised learning (SSL) partially circumvents the high cost of\nlabeling data by augmenting a small labeled dataset with a large and relatively\ncheap unlabeled dataset drawn from the same distribution. This paper offers a\nnovel interpretation of two deep learning-based SSL approaches, ladder networks\nand virtual adversarial training (VAT), as applying distributional smoothing to\ntheir respective latent spaces. We propose a class of models that fuse these\napproaches. We achieve near-supervised accuracy with high consistency on the\nMNIST dataset using just 5 labels per class: our best model, ladder with\nlayer-wise virtual adversarial noise (LVAN-LW), achieves 1.42% +/- 0.12 average\nerror rate on the MNIST test set, in comparison with 1.62% +/- 0.65 reported\nfor the ladder network. On adversarial examples generated with L2-normalized\nfast gradient method, LVAN-LW trained with 5 examples per class achieves\naverage error rate 2.4% +/- 0.3 compared to 68.6% +/- 6.5 for the ladder\nnetwork and 9.9% +/- 7.5 for VAT.","url_abs":"http://arxiv.org/abs/1711.07476v2","url_pdf":"http://arxiv.org/pdf/1711.07476v2.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":"virtual-adversarial-ladder-networks-for-semi","repo_url":"https://github.com/sakishinoda/tf-ssl","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null},{"paper_slug":"virtual-adversarial-ladder-networks-for-semi","repo_url":"https://github.com/yyll008/yyll008.github.io","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}