Differential Privacy for Clustering Under Continual Observation

Abstract
We consider the problem of clustering privately a dataset in that undergoes both insertion and deletion of points. Specifically, we give an -differentially private clustering mechanism for the -means objective under continual observation. This is the first approximation algorithm for that problem with an additive error that depends only logarithmically in the number of updates. The multiplicative error is almost the same as non privately. To do so we show how to perform dimension reduction under continual observation and combine it with a differentially private greedy approximation algorithm for -means. We also partially extend our results to the -median problem.
View on arXivComments on this paper