Jacob Abernethy : [email protected]
Note : Mail sparingly; Piazza is the preferred mode of communication.

• Zihao Hu
• Yeojoon Youn
• Guanghui Wang
• Tyler LaBonte

TBA

Weber SST III, Lecture Hall 1

• Advanced Linear Algebra
• Probability and Statistics
• Convex Optimization/Analysis

Foundations of Machine Learning by Mehyrar Mohri et. al (FML)

1. Basics - 4 lectures
   • Basic Linear Algebra
   • Probability
   • Inequalities
   • Reference : Appendices of FML
2. Statistical Learning Theory - 7 lectures
   • What is generalization?
   • Relies on IID (Independent and Identically Distributed) assumption
   • We try to answer questions like, Why training error goes down but test error goes up?
   • Tools : VC Dimension, Rademacher complexity, Chernoff bounds, Union bounds
3. Online Learning & Bandits Problem - 9 lectures
   • Non-statistical approach to learning
   • Regret minimization
   • Useful for game theory settings
   • Useful for optimization
4. Extracurricular - 4 lectures
   • RL Theory
   • Differential Privacy
   • Sampling

1. Homeworks : 40%
2. Exam : 30% (30th March, 2023)
3. Final project : 20%
4. Participation (Scribes) : 10%

TBA; Professor mentioned a potential one week HW late policy (distributed throughout the semester)

• Class is intended to follow a 35-5-35 minute split to break up content with a 5 minute break.
• Some TAs will be delivering lectures pertaining to their research experience and background.