{"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/11125745","title":"Bayesian Active Learning for Classification and Preference Learning","arxiv_id":"1112.5745","date":"2011-12-24","proceeding":null,"authors":["Neil Houlsby","Ferenc Huszár","Zoubin Ghahramani","Máté Lengyel"],"abstract":"Information theoretic active learning has been widely studied for\nprobabilistic models. For simple regression an optimal myopic policy is easily\ntractable. However, for other tasks and with more complex models, such as\nclassification with nonparametric models, the optimal solution is harder to\ncompute. Current approaches make approximations to achieve tractability. We\npropose an approach that expresses information gain in terms of predictive\nentropies, and apply this method to the Gaussian Process Classifier (GPC). Our\napproach makes minimal approximations to the full information theoretic\nobjective. Our experimental performance compares favourably to many popular\nactive learning algorithms, and has equal or lower computational complexity. We\ncompare well to decision theoretic approaches also, which are privy to more\ninformation and require much more computational time. Secondly, by developing\nfurther a reformulation of binary preference learning to a classification\nproblem, we extend our algorithm to Gaussian Process preference learning.","url_abs":"http://arxiv.org/abs/1112.5745v1","url_pdf":"http://arxiv.org/pdf/1112.5745v1.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":"11125745","repo_url":"https://github.com/airi-institute/al_toolbox","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"11125745","repo_url":"https://github.com/cambridge-mlg/BALaudiogram","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"active-learning","task_name":"Active Learning"},{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"}],"methods":[{"method_slug":"gaussian-process","method_name":"Gaussian Process"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1112.5745","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}