Abstract Number:
1048
Submission Type:
Invited Paper Session
Participants:
Bei Jiang (1), Naisyin Wang (2), Bei Jiang (1), Christian Robert (3), Adam Smith (4), Eric Kolaczyk (5)
Institutions:
(1) University of Alberta, Edmonton, Alberta, Canada, (2) University of Michigan, Ann Arbor, Michigan, USA, (3) Universite Paris Dauphine, Paris, France, (4) Boston University, Boston, Massachusetts, USA, (5) McGill University, Montreal, Quebec, Canada
Chair:
Discussant:
Session Organizer:
Speaker(s):
Session Description:
Differential privacy, a mathematically rigorous framework for privacy preservation, has played a key role in driving progress on this front. Despite that differential privacy is rooted in Statistics and is intended for use in statistical data analysis, much of the research on this topic has been conducted by computer scientists. However, the way that the two research communities formulate and tackle problems differ profoundly. This session will provide a platform for leading statisticians and computer scientists working on privacy to share the insights on the latest developments in their respective fields, exchange ideas, and identify key areas for future collaboration.
The session will consist of three oral presentations by experts from both differential privacy and statistical analysis backgrounds, focusing on fundamental concepts of differential privacy and statistical analysis and highlighting their relevance in advancing privacy-preserving data analysis. These research talks will provide deep dives into specific methodologies, real-world implementations, and emerging trends. The session will be then followed by an engaging discussion led by a discussant with the goal to identify topics that are at the intersection of statistics and computing science, and to identify models and analysis that are different in the different domains. These discussions will foster engaging dialogues on challenges, opportunities, and potential collaborations between the two fields.
The invited session will appeal to a broad spectrum of participants, including researchers, academics, practitioners, and policymakers. Professionals engaged in differential privacy, statistical analysis, and related fields will find common ground to explore novel approaches, gain insights into cutting-edge research, and forge connections with like-minded experts. The session's interdisciplinary nature caters to those passionate about advancing privacy-preserving data analysis across domains.
The session will consist of speakers from diverse backgrounds, including both male and female experts from different regions (Canada, US and France/UK), which ensures a comprehensive coverage of topics, enriching discussions through multifaceted viewpoints.
Specifically, this session will cover the following presentations:
Speaker 1:
Christian Robert, Professor of Statistics
Université Paris-Dauphine, Paris, France
University of Warwick, Coventry, UK
Talk title: Bayesian hierarchical modelling for data privacy and federated learning
Speaker 2:
Adam Smith, Professor of Computer Science and Electrical and Computer Engineering
Boston University, Boston, Massachusetts, USA
Talk title: Fast algorithms for differentially private statistical estimation
Speaker 3:
Eric Kolaczyk, Professor of Statistics
McGill University, Montreal, Quebec, Canada
Talk title: Differentially private linear regression with linked data
Discussant:
Naisyin Wang, Professor of Statistics
University of Michigan, Ann Arbor, Michigan, USA
Sponsors:
Caucus for Women in Statistics 3
SSC (Statistical Society of Canada) 1
WNAR 2
Theme:
Statistics and Data Science: Informing Policy and Countering Misinformation
Yes
Applied
Yes
Estimated Audience Size
Medium (80-150)
I have read and understand that JSM participants must abide by the Participant Guidelines.
Yes
I understand and have communicated to my proposed speakers that JSM participants must register and pay the appropriate registration fee by June 1, 2024. The registration fee is nonrefundable.
I understand