Individual Privacy Accounting via a Renyi Filter
This work addresses the need for more efficient privacy accounting in adaptive data analysis, particularly for typical data points, though it is incremental as it builds on existing filter methods.
The paper tackles the problem of overly conservative privacy loss accounting in sequential data analyses by introducing a method for tighter, personalized privacy loss estimates per individual, resulting in a simpler and more practical filter for Rényi differential privacy that improves the privacy-utility tradeoff.
We consider a sequential setting in which a single dataset of individuals is used to perform adaptively-chosen analyses, while ensuring that the differential privacy loss of each participant does not exceed a pre-specified privacy budget. The standard approach to this problem relies on bounding a worst-case estimate of the privacy loss over all individuals and all possible values of their data, for every single analysis. Yet, in many scenarios this approach is overly conservative, especially for "typical" data points which incur little privacy loss by participation in most of the analyses. In this work, we give a method for tighter privacy loss accounting based on the value of a personalized privacy loss estimate for each individual in each analysis. To implement the accounting method we design a filter for Rényi differential privacy. A filter is a tool that ensures that the privacy parameter of a composed sequence of algorithms with adaptively-chosen privacy parameters does not exceed a pre-specified budget. Our filter is simpler and tighter than the known filter for $(ε,δ)$-differential privacy by Rogers et al. We apply our results to the analysis of noisy gradient descent and show that personalized accounting can be practical, easy to implement, and can only make the privacy-utility tradeoff tighter.