Privacy Model

Privacy models aim to safeguard sensitive data used in machine learning and data analysis while preserving data utility. Current research focuses on developing and auditing differentially private algorithms, including those for clustering, and exploring alternative models like shuffle privacy and local differential privacy, with a strong emphasis on mitigating privacy risks in federated learning and other distributed settings. These advancements are crucial for enabling responsible data utilization in various applications, balancing the need for data-driven insights with robust privacy protections.