CS 700
Spring 2026
Doctoral
Seminar in Trustworthy AI
A research seminar examining fairness, privacy, and explainability in modern machine learning systems through primary literature.
Course overview
This doctoral seminar is a close reading of primary literature on fairness, privacy, and explainability in machine learning. Each week, students lead discussion on a small cluster of papers and are expected to identify open problems suitable for a course research project or thesis chapter.
Topics covered
- Formal fairness definitions and their trade-offs
- Differential privacy: definitions, mechanisms, and composition
- Explainability: post-hoc methods versus explainable-by-design approaches
- Robustness and adversarial considerations in deployed ML systems
- Emerging regulatory and policy context for trustworthy AI
Prerequisites
A graduate-level machine learning course (e.g., CS 501) and instructor permission. This seminar assumes comfort reading and critiquing research papers independently.
Grading
- 40% — Weekly discussion leadership and participation
- 30% — Independent research project or literature review
- 30% — Final presentation to the seminar