On the Protection of Private Information in Machine Learning Systems: Two Recent Approaches
This work provides a comparative analysis for researchers and practitioners interested in privacy protection in ML, but it is incremental as it reviews existing approaches without introducing new methods.
The paper reviews two recent works on privacy in machine learning systems by comparing them with early privacy principles from the 1970s, highlighting that older ideas may still be valid and useful in addressing data privacy concerns.
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.