Hannah Keller

Title: Innovations by Inventor Hannah Keller

Introduction

Hannah Keller is a prominent inventor based in Mannheim, Germany. She has made significant contributions to the field of deep learning, particularly in the area of privacy and data security. Her innovative work focuses on developing frameworks that enhance the interpretability of differentially private algorithms.

Latest Patents

Hannah Keller holds a patent for an "Interpretability framework for differentially private deep learning." This patent addresses the challenge of re-identifying data points from a dataset based on a differentially private function output. The framework specifies a bound for an adversarial posterior belief ρ, which corresponds to the likelihood of re-identification. Privacy parameters ε and δ are calculated based on the received data, governing the differential privacy (DP) algorithm applied to a function evaluated over the dataset. The calculations are based on the ratio of probability distributions of different observations, bound by the posterior belief ρ as applied to the dataset. The calculated privacy parameters are then utilized to implement the DP algorithm effectively.

Career Highlights

Hannah Keller is currently employed at SAP SE, where she continues to innovate and contribute to advancements in technology. Her work is pivotal in ensuring that deep learning models can operate securely while maintaining user privacy.

Collaborations

Hannah collaborates with notable colleagues, including Daniel Bernau and Philip-William Grassal. Their combined expertise fosters a dynamic environment for innovation and research.

Conclusion

Hannah Keller's contributions to the field of deep learning and privacy are noteworthy. Her patent and ongoing work at SAP SE highlight her commitment to advancing technology while prioritizing data security.