{"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/a-differentially-private-kernel-two-sample","title":"A Differentially Private Kernel Two-Sample Test","arxiv_id":"1808.00380","date":"2018-08-01","proceeding":null,"authors":["Anant Raj","Ho Chung Leon Law","Dino Sejdinovic","Mijung Park"],"abstract":"Kernel two-sample testing is a useful statistical tool in determining whether\ndata samples arise from different distributions without imposing any parametric\nassumptions on those distributions. However, raw data samples can expose\nsensitive information about individuals who participate in scientific studies,\nwhich makes the current tests vulnerable to privacy breaches. Hence, we design\na new framework for kernel two-sample testing conforming to differential\nprivacy constraints, in order to guarantee the privacy of subjects in the data.\nUnlike existing differentially private parametric tests that simply add noise\nto data, kernel-based testing imposes a challenge due to a complex dependence\nof test statistics on the raw data, as these statistics correspond to\nestimators of distances between representations of probability measures in\nHilbert spaces. Our approach considers finite dimensional approximations to\nthose representations. As a result, a simple chi-squared test is obtained,\nwhere a test statistic depends on a mean and covariance of empirical\ndifferences between the samples, which we perturb for a privacy guarantee. We\ninvestigate the utility of our framework in two realistic settings and conclude\nthat our method requires only a relatively modest increase in sample size to\nachieve a similar level of power to the non-private tests in both settings.","url_abs":"http://arxiv.org/abs/1808.00380v1","url_pdf":"http://arxiv.org/pdf/1808.00380v1.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":"a-differentially-private-kernel-two-sample","repo_url":"https://github.com/antoninschrab/dpkernel-paper","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"jax","reach":null},{"paper_slug":"a-differentially-private-kernel-two-sample","repo_url":"https://github.com/hcllaw/private_tst","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"hypothesis-testing","task_name":"Two-sample testing"},{"task_slug":"two","task_name":"Vocal Bursts Valence Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}