STATS 700, Fall 2025
This half semester course will cover some “recent hits” in learning theory. After quickly discussing some background material in classical learning theory, both statistical and online, we will look at some recent progress (even some breakthroughs!) in fundamental problems like:
- sample compression schemes
- multiclass classification
- private learning
- operator learning
- learning from quantum examples
- generation
No prior background in learning theory is required. But a high level of mathematical maturity will be needed to fully benefit from this course. The topics list below is tentative and subject to change.
Logistics
Time & Days: TuTh 2:30PM - 4:00PM
Location: TBD
Half semester course dates: Aug 25, 2025-Oct 10, 2025
Topics
Background
- Understanding Machine Learning
- Part 1 (Foundations)
- Basics of statistical learning theory up until Chapter 6 (the VC-Dimension)
- Part 3 (Additional Learning Models)
- Chapter 21 (Online Learning)
- Part 4 (Advanced Theory)
- Chapter 29 (Multiclass Learnability)
- Chapter 30 (Compression Bounds)
Sample Compression Schemes
Multiclass classification
Private learning
Operator learning
Learning with quantum examples
Generation