{"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/mimic-and-classify-a-meta-algorithm-for","title":"Mimic and Classify : A meta-algorithm for Conditional Independence Testing","arxiv_id":"1806.09708","date":"2018-06-25","proceeding":null,"authors":["Rajat Sen","Karthikeyan Shanmugam","Himanshu Asnani","Arman Rahimzamani","Sreeram Kannan"],"abstract":"Given independent samples generated from the joint distribution\n$p(\\mathbf{x},\\mathbf{y},\\mathbf{z})$, we study the problem of Conditional\nIndependence (CI-Testing), i.e., whether the joint equals the CI distribution\n$p^{CI}(\\mathbf{x},\\mathbf{y},\\mathbf{z})= p(\\mathbf{z})\np(\\mathbf{y}|\\mathbf{z})p(\\mathbf{x}|\\mathbf{z})$ or not. We cast this problem\nunder the purview of the proposed, provable meta-algorithm, \"Mimic and\nClassify\", which is realized in two-steps: (a) Mimic the CI distribution close\nenough to recover the support, and (b) Classify to distinguish the joint and\nthe CI distribution. Thus, as long as we have a good generative model and a\ngood classifier, we potentially have a sound CI Tester. With this modular\nparadigm, CI Testing becomes amiable to be handled by state-of-the-art, both\ngenerative and classification methods from the modern advances in Deep\nLearning, which in general can handle issues related to curse of dimensionality\nand operation in small sample regime. We show intensive numerical experiments\non synthetic and real datasets where new mimic methods such conditional GANs,\nRegression with Neural Nets, outperform the current best CI Testing performance\nin the literature. Our theoretical results provide analysis on the estimation\nof null distribution as well as allow for general measures, i.e., when either\nsome of the random variables are discrete and some are continuous or when one\nor more of them are discrete-continuous mixtures.","url_abs":"http://arxiv.org/abs/1806.09708v1","url_pdf":"http://arxiv.org/pdf/1806.09708v1.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":"mimic-and-classify-a-meta-algorithm-for","repo_url":"https://github.com/rajatsen91/mimic_classify","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}