EN.601.774 Theory of Replicable ML

Course Info

Lecture: Tue/Thu 3:00 - 4:15pm, Hodson 316

Instructor: Jess Sorrell (jess@jhu.edu), Malone 303

TA: Iliana Maifeld-Carucci (imaifel1@jhu.edu)

Office Hours: Jess - Friday 12pm (though if no one shows up in the first 15 min, I may leave) or by appointment
Iliana - by appointment

Syllabus: https://jess-sorrell.github.io/Courses/Replicable-ML/syllabus.pdf

Course Description

Replicability is vital to ensuring scientific conclusions are reliable, but failures of replicability have been a major issue in nearly all scientific areas of study, and machine learning is no exception. In this course, we will study replicability as a property of learning and other statistical algorithms, developing a theory of replicable learning. We will cover recent formalizations of replicability and their relationships to other common stability notions such as differential privacy and adaptive generalization. We will survey replicable algorithms for fundamental learning tasks, and discuss the limitations of replicable algorithms. If time permits, we will discuss replicability in other settings, such as reinforcement learning and clustering, or other useful and related stability notions such as list replicability and global stability.

Lecture Notes

Project Ideas

Resources