{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2023,1,13]],"date-time":"2023-01-13T06:12:30Z","timestamp":1673590350171},"reference-count":0,"publisher":"Privacy Enhancing Technologies Symposium Advisory Board","issue":"1","license":[{"start":{"date-parts":[[2023,1,1]],"date-time":"2023-01-01T00:00:00Z","timestamp":1672531200000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["PoPETs"],"abstract":"<jats:p>Private multi-winner voting is the task of revealing k-hot binary vectors satisfying a bounded differential privacy (DP) guarantee. This task has been understudied in machine learning literature despite its prevalence in many domains such as healthcare. We propose three new DP multi-winner mechanisms: Binary, Tau, and Powerset voting. Binary voting operates independently per label through composition. Tau voting bounds votes optimally in their L2 norm for tight data-independent guarantees. Powerset voting operates over the entire binary vector by viewing the possible outcomes as a power set. Our theoretical and empirical analysis shows that Binary voting can be a competitive mechanism on many tasks unless there are strong correlations between labels, in which case Powerset voting outperforms it. We use our mechanisms to enable privacy-preserving multi-label learning in the central setting by extending the canonical single-label technique: PATE. We find that our techniques outperform current state-of-the-art approaches on large, real-world healthcare data and standard multi-label benchmarks. We further enable multi-label confidential and private collaborative (CaPC) learning and show that model performance can be significantly improved in the multi-site setting.<\/jats:p>","DOI":"10.56553\/popets-2023-0031","type":"journal-article","created":{"date-parts":[[2023,1,12]],"date-time":"2023-01-12T16:43:52Z","timestamp":1673541832000},"page":"527-555","source":"Crossref","is-referenced-by-count":0,"title":["Private Multi-Winner Voting for Machine Learning"],"prefix":"10.56553","volume":"2023","author":[{"given":"Adam","family":"Dziedzic","sequence":"first","affiliation":[{"name":"University of Toronto and Vector Institute"}]},{"given":"Christopher","family":"A. Choquette-Choo","sequence":"additional","affiliation":[{"name":"Google Research, Brain Tea"}]},{"given":"Natalie","family":"Dullerud","sequence":"additional","affiliation":[{"name":"University of Toronto and Vector Institute"}]},{"given":"Vinith","family":"Suriyakumar","sequence":"additional","affiliation":[{"name":"MIT"}]},{"given":"Ali","family":"Shahin Shamsabadi","sequence":"additional","affiliation":[{"name":"The Alan Turing Institue"}]},{"given":"Muhammad","family":"Ahmad Kaleem","sequence":"additional","affiliation":[{"name":"University of Toronto and Vector Institute"}]},{"given":"Somesh","family":"Jha","sequence":"additional","affiliation":[{"name":"University of Wisconsin-Madison"}]},{"given":"Nicolas","family":"Papernot","sequence":"additional","affiliation":[{"name":"University of Toronto and Vector Institute"}]},{"given":"Xiao","family":"Wang","sequence":"additional","affiliation":[{"name":"Northwestern University"}]}],"member":"35752","published-online":{"date-parts":[[2023,1]]},"container-title":["Proceedings on Privacy Enhancing Technologies"],"original-title":[],"deposited":{"date-parts":[[2023,1,12]],"date-time":"2023-01-12T16:44:25Z","timestamp":1673541865000},"score":1,"resource":{"primary":{"URL":"https:\/\/petsymposium.org\/popets\/2023\/popets-2023-0031.php"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,1]]},"references-count":0,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2023,1]]}},"alternative-id":["10.56553\/popets-2023-0031"],"URL":"https:\/\/doi.org\/10.56553\/popets-2023-0031","relation":{},"ISSN":["2299-0984"],"issn-type":[{"value":"2299-0984","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,1]]}}}