Chasing Your Long Tails - Differentially Private Prediction in Health Care Settings

This course teaches differentially private prediction in health care settings. The learning outcomes include understanding the concept of differential privacy, exploring the unique aspects of machine learning in health care, and learning about datasets and differentially private training. The course covers extreme tradeoffs in health care, group fairness defined by influence, important considerations, and future directions. The intended audience for this course is individuals interested in privacy-preserving machine learning in health care. The teaching method is through a research presentation at the FAccT 2021 conference.