%0 Journal Article %1 nguyen2019efficient %A Nguyen, Huy L. %A Ullman, Jonathan %A Zakynthinou, Lydia %D 2019 %K convergence differential-privacy %T Efficient Private Algorithms for Learning Halfspaces %U http://arxiv.org/abs/1902.09009 %X We present new differentially private algorithms for learning a large-margin halfspace. In contrast to previous algorithms, which are based on either differentially private simulations of the statistical query model or on private convex optimization, the sample complexity of our algorithms depends only on the margin of the data, and not on the dimension.

@article{nguyen2019efficient, abstract = {We present new differentially private algorithms for learning a large-margin halfspace. In contrast to previous algorithms, which are based on either differentially private simulations of the statistical query model or on private convex optimization, the sample complexity of our algorithms depends only on the margin of the data, and not on the dimension.}, added-at = {2020-01-29T19:19:57.000+0100}, author = {Nguyen, Huy L. and Ullman, Jonathan and Zakynthinou, Lydia}, biburl = {https://www.bibsonomy.org/bibtex/2513312b49f94ea6f19965a7b4e8d27db/kirk86}, description = {[1902.09009] Efficient Private Algorithms for Learning Halfspaces}, interhash = {881a4f828b55300c0eface3b6e05f4d0}, intrahash = {513312b49f94ea6f19965a7b4e8d27db}, keywords = {convergence differential-privacy}, note = {cite arxiv:1902.09009Comment: 21 pages}, timestamp = {2020-01-29T19:19:57.000+0100}, title = {Efficient Private Algorithms for Learning Halfspaces}, url = {http://arxiv.org/abs/1902.09009}, year = 2019 }