DC Field | Value | Language |
---|---|---|
dc.contributor.author | Lee, Kangwook | ko |
dc.contributor.author | Kim, Hoon | ko |
dc.contributor.author | Lee, Kyoungmin | ko |
dc.contributor.author | Suh, Changho | ko |
dc.contributor.author | Ramchandran, Kannan | ko |
dc.date.accessioned | 2019-12-13T10:27:10Z | - |
dc.date.available | 2019-12-13T10:27:10Z | - |
dc.date.created | 2019-11-26 | - |
dc.date.created | 2019-11-26 | - |
dc.date.created | 2019-11-26 | - |
dc.date.created | 2019-11-26 | - |
dc.date.issued | 2019-07-05 | - |
dc.identifier.citation | Proceedings of the IEEE International Symposium on Information Theory(ISIT), pp.542 - 546 | - |
dc.identifier.uri | http://hdl.handle.net/10203/269354 | - |
dc.description.abstract | The goal of differentially private data publishing is to release a modified dataset so that its privacy can be ensured while allowing for efficient learning. We propose a new data publishing algorithm in which a released dataset is formed by mixing l randomly chosen data points and then perturbing them with an additive noise. Our privacy analysis shows that as l increases, noise with smaller variance is sufficient to achieve a target privacy level. In order to quantify the usefulness of our algorithm, we adopt the accuracy of a predictive model trained with our synthetic dataset, which we call the utility of the dataset. By characterizing the utility of our dataset as a function of l, we show that one can learn both linear and nonlinear predictive models so that they yield reasonably good prediction accuracies. Particularly, we show that there exists a sweet spot on l that maximizes the prediction accuracy given a required privacy level, or vice versa. We also demonstrate that given a target privacy level, our datasets can achieve higher utility than other datasets generated with the existing data publishing algorithms. | - |
dc.language | English | - |
dc.publisher | Proceedings of the IEEE International Symposium on Information Theory | - |
dc.title | Synthesizing differentially private datasets using random mixing | - |
dc.type | Conference | - |
dc.identifier.wosid | 000489100300110 | - |
dc.identifier.scopusid | 2-s2.0-85073169963 | - |
dc.type.rims | CONF | - |
dc.citation.beginningpage | 542 | - |
dc.citation.endingpage | 546 | - |
dc.citation.publicationname | Proceedings of the IEEE International Symposium on Information Theory(ISIT) | - |
dc.identifier.conferencecountry | FR | - |
dc.identifier.conferencelocation | Paris, France | - |
dc.identifier.doi | 10.1109/ISIT.2019.8849381 | - |
dc.contributor.localauthor | Suh, Changho | - |
dc.contributor.nonIdAuthor | Lee, Kangwook | - |
dc.contributor.nonIdAuthor | Kim, Hoon | - |
dc.contributor.nonIdAuthor | Lee, Kyoungmin | - |
dc.contributor.nonIdAuthor | Ramchandran, Kannan | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.