CMPS 290C: Advanced Machine Learning: Responsible Data Science

Time: Wed 1:20-4:20

Location: E2 506

Instructor: Lise Getoor

Office Hours: Tue 3:30-5:30 E2 341B (Updated!)


Machine learning and data science are enjoying an explosion of interest, across all areas of endeavor: business, academia, and government are all embracing new data-driven approaches to utilize the wide range of data with which we are now inundated.  With the growing popularity of machine learning and data-driven methods also comes concerns. These concerns range from basic statistical and computational literacy (understanding key concepts such as the curse of dimensionality, overfitting, biased samples, computational complexity, data ownership, security, etc.) to emerging topics such as fairness, accountability and interpretability in machine learning.   This course will serve as an introduction to these topics. The course will begin with basic introductory material on key topics such as: privacy, causality, interpretability, fairness and reproducibility, and the latter part of the course will include readings based on the class interests. Students should come away with a modern grounding in emerging topics of ethics and responsibility in data science.


Course requirements: Class attendance and participation, a quarter long project which can either be a literature survey or a research project (students are highly encouraged to choose a topic which aligns with their MS or PhD research), and discussion/presentation of research papers.


Prerequisites: basic background in machine learning and statistics recommended but not required.


Permission Code: Please contact the instructor if you need a permission code,