FALL 2023

IDEAL Special Program on Trustworthy and Reliable Data Science

Massive datasets are susceptible to various kinds of noise and corruptions. Many data analysis primitives are brittle to even small corruptions of the data sets, while modern sophisticated machine learning systems, despite having human-level performance at various tasks, do not have (anywhere near) human-level robustness. As data science and machine learning are deployed in almost every aspect of decision-making, it is vital to understand when and how we can design methods and systems that are provably reliable and trustworthy.

This special program aims to bring together researchers from different disciplines to explore methods and algorithms for data science that are reliable and trustworthy under various settings like 1) failure of model assumptions due to contamination or modeling errors, (2) adversarial behavior in the system, (3) distribution shift from natural variations in data, (4) distributed settings with unreliable agents like in federated learning.

Organizers

• Saba Ahmadi (Toyota Technological Institute of Chicago)
• Avrim Blum (Toyota Technological Institute of Chicago)
• Chao Gao (University of Chicago)
• Varun Gupta (University of Chicago)
• Daniel Linna (Northwestern University)
• Mesrob Ohanessian (University of Illinois Chicago)
• Liren Shan (Northwestern University)
• Gyuri Turan (University of Illinois at Chicago)
• Aravindan Vijayaraghavan (Northwestern University)
• Binghui Wang (Illinois Institute of Technology)
• Ren Wang (Illinois Institute of Technology)
• Haifeng Xu (University of Chicago)

Graduate Courses

click here to visit IDEAL Fall 2023 Course Catalog 

Weekly Reading Group

• Topic: Learning with Untrusted Data
• Times: Tuesdays, 4-5PM (First meeting on September 19th)
• Location: Virtual
• Organizer: Liren Shan
• Register: Click here to register