Martin Dunsche
CASA – Cluster of Excellence for Cyber Security, PhD Candidate at Chair of Stochastics, Ruhr University Bochum
9th November 2023, 4:00pm - 5:00pm (GST)
| Title: | Differential Privacy – how private is your algorithm? |
| Abstract: | Differential privacy is a fundamental concept in data privacy and security. It provides a rigorous framework for protecting sensitive information while still allowing valuable insights to be drawn from data. Assessing the privacy guarantees of an algorithm is therefore a fundamental task in Differential Privacy. In this paper, we present a new estimator in a black-box scenario for the Rényi divergence, which is transformed into a statistical lower bound. |
| Bio: | 2018- March 2021 Master of Science Mathematics at the Ruhr-University Bochum May 2021- PhD candidate at the Chair of Stochastics May 2021- CASA PhD candidate – Differential Privacy |