Perturbation of convex risk minimization and its application in differential private learning algorithms.

Perturbation of convex risk minimization and its application in differential private learning algorithms.
复制标题

凸风险最小化扰动及其在差分隐私学习算法中的应用

DOI:
10.1186/s13660-016-1280-0
复制
发表时间:
2017
影响因子:
1.6
通讯作者:
Wang C
Wang C
中科院分区:
数学3区
文献类型:
--
作者:
Nie W;Wang C

文献摘要

参考文献

相似文献

凸风险最小化是学习理论中常用的设置。本文首先对这类算法进行了扰动分析,然后将这一结果应用于差分私人学习算法。我们的分析要求目标函数是强凸的。这使得我们在构造差分私有算法时,将前面的分析扩展到不可微损失函数。最后,进行了误差分析,说明了参数的选择。
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.
DOI: 10.2307/2282952
发表时间: 1963-01-01
影响因子: 3.7
作者:
HOEFFDING, W
通讯作者: HOEFFDING, W
DOI: 10.3150/10-bej267
发表时间: 2011-02-01
期刊: BERNOULLI
影响因子: 1.5
作者:
Steinwart, Ingo;Christmann, Andreas
通讯作者: Christmann, Andreas
DOI: 10.1016/j.jco.2010.10.002
发表时间: 2011-02-01
影响因子: 1.7
作者:
Wang, Cheng;Zhou, Ding-Xuan
通讯作者: Zhou, Ding-Xuan
具有不同高斯的条件分位数
DOI: 10.1007/s10444-011-9257-5
发表时间: 2013-05-01
影响因子: 1.7
作者:
Xiang, Dao-Hong
通讯作者: Xiang, Dao-Hong
DOI: 10.1162/153244302760200704
发表时间: 2002-06-01
影响因子: 6
作者:
Bousquet, O;Elisseeff, A
通讯作者: Elisseeff, A