Malhar Jere
PhD student at UCSD. Specializing in Privacy Preserving Machine Learning
Highlighted by Malhar Jere
From A Brief Introduction to Differential Privacy by Georgian Partners
How does differential privacy work if the privacy loss grows so quickly? To ensure a meaningful privacy guarantee, data curators can enforce a maximum privacy loss. If the number of queries exceeds the threshold, then the privacy guarantee becomes too weak and the curator stops answering queries. The maximum privacy loss is called the privacy budget [11]. We can think of each query as a privac…