摘要

        In the rapidly advancing field of data analytics, the dual imperatives of maintaining privacy while ensuring utility present significant challenges. This talk explores the intricate dynamics of data privacy in the context of burgeoning machine learning applications. It examines the ongoing debates and the ad-hoc nature of privacy definitions, techniques, and regulations, underscoring the discrepancies between legal frameworks and their practical application. The discussion centers on pioneering research methodologies that aim to surpass the conventional trade-offs between privacy and utility. A significant emphasis is placed on collaborative genome-wide association studies (GWAS), which serve as a primary case study to demonstrate these innovative approaches. These examples highlight the application of new strategies in protecting data privacy without compromising the utility essential for scientific progress.

 个人简介 

       王欣月，现任中国人民大学统计学数据科学与大数据统计系讲师，专注于数据安全与隐私研究。在重庆大学取得统计学学士学位后，于Rutgers University先后获得经济学硕士和管理学(信息技术方向)博士学位。研究主要围绕数据分析、隐私保护及安全性问题，尤其是在开发和改进安全的机器学习与人工智能算法方面。此外，还探索这些算法在金融和生物信息学等领域的应用，推动技术的前沿发展和实际应用。