On the Protection of Private Information in Machine Learning Systems: Two Recent Approaches

Martín Abadi,Úlfar Erlingsson,Ian Goodfellow,H. Brendan McMahan,Ilya Mironov,Nicolas Papernot,Kunal Talwar,Li Zhang
DOI: https://doi.org/10.48550/arXiv.1708.08022
IF: 5.414
2017-08-26
Machine Learning
Abstract:The recent, remarkable growth of machine learning has led to intense interest in the privacy of the data on which machine learning relies, and to new techniques for preserving privacy. However, older ideas about privacy may well remain valid and useful. This note reviews two recent works on privacy in the light of the wisdom of some of the early literature, in particular the principles distilled by Saltzer and Schroeder in the 1970s.
What problem does this paper attempt to address?