Perturbation of convex risk minimization and its application in differential private learning algorithms
Abstract Convex risk minimization is a commonly used setting in learning theory. In this paper, we firstly give a perturbation analysis for such algorithms, and then we apply this result to differential private learning algorithms. Our analysis needs the objective functions to be strongly convex. Th...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2017-01-01
|
| Series: | Journal of Inequalities and Applications |
| Subjects: | |
| Online Access: | http://link.springer.com/article/10.1186/s13660-016-1280-0 |
| _version_ | 1811259203648487424 |
|---|---|
| author | Weilin Nie Cheng Wang |
| author_facet | Weilin Nie Cheng Wang |
| author_sort | Weilin Nie |
| collection | DOAJ |
| description | Abstract Convex risk minimization is a commonly used setting in learning theory. In this paper, we firstly give a perturbation analysis for such algorithms, and then we apply this result to differential private learning algorithms. Our analysis needs the objective functions to be strongly convex. This leads to an extension of our previous analysis to the non-differentiable loss functions, when constructing differential private algorithms. Finally, an error analysis is then provided to show the selection for the parameters. |
| first_indexed | 2024-04-12T18:26:58Z |
| format | Article |
| id | doaj.art-09cdec8de9694e40b1517767b567728a |
| institution | Directory Open Access Journal |
| issn | 1029-242X |
| language | English |
| last_indexed | 2024-04-12T18:26:58Z |
| publishDate | 2017-01-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | Journal of Inequalities and Applications |
| spelling | doaj.art-09cdec8de9694e40b1517767b567728a2022-12-22T03:21:12ZengSpringerOpenJournal of Inequalities and Applications1029-242X2017-01-012017111610.1186/s13660-016-1280-0Perturbation of convex risk minimization and its application in differential private learning algorithmsWeilin Nie0Cheng Wang1Huizhou UniversityHuizhou UniversityAbstract Convex risk minimization is a commonly used setting in learning theory. In this paper, we firstly give a perturbation analysis for such algorithms, and then we apply this result to differential private learning algorithms. Our analysis needs the objective functions to be strongly convex. This leads to an extension of our previous analysis to the non-differentiable loss functions, when constructing differential private algorithms. Finally, an error analysis is then provided to show the selection for the parameters.http://link.springer.com/article/10.1186/s13660-016-1280-0differential privacyconvex risk minimizationperturbationconcentration inequalityerror decomposition |
| spellingShingle | Weilin Nie Cheng Wang Perturbation of convex risk minimization and its application in differential private learning algorithms Journal of Inequalities and Applications differential privacy convex risk minimization perturbation concentration inequality error decomposition |
| title | Perturbation of convex risk minimization and its application in differential private learning algorithms |
| title_full | Perturbation of convex risk minimization and its application in differential private learning algorithms |
| title_fullStr | Perturbation of convex risk minimization and its application in differential private learning algorithms |
| title_full_unstemmed | Perturbation of convex risk minimization and its application in differential private learning algorithms |
| title_short | Perturbation of convex risk minimization and its application in differential private learning algorithms |
| title_sort | perturbation of convex risk minimization and its application in differential private learning algorithms |
| topic | differential privacy convex risk minimization perturbation concentration inequality error decomposition |
| url | http://link.springer.com/article/10.1186/s13660-016-1280-0 |
| work_keys_str_mv | AT weilinnie perturbationofconvexriskminimizationanditsapplicationindifferentialprivatelearningalgorithms AT chengwang perturbationofconvexriskminimizationanditsapplicationindifferentialprivatelearningalgorithms |