ECE 532 | Matrix Methods in Machine Learning

Syllabus

DateProgramInstructor
Course introduction
Grigorios Chrysos
––
Vectors
Grigorios Chrysos
––
Vectors and Matrices
Linear Systems, Norms and Projections
Grigorios Chrysos
–
Least Squares
Grigorios Chrysos
–––
The Singular Value Decomposition
Grigorios Chrysos
First midterm
Grigorios Chrysos
–
Applications of the SVD
Grigorios Chrysos
–
PageRank and the SVD
Grigorios Chrysos
Loss functions
Omar Chehab
Optimization algorithms
Omar Chehab
Convergence assumptions
Homework assignment
Omar Chehab
Convergence guarantees
Omar Chehab
Second midterm
Omar Chehab
TBD
Homework assignment
Omar Chehab
TBD
Quiz
Omar Chehab
TBD
Omar Chehab
TBD
Quiz
Omar Chehab
TBD
Omar Chehab
Revision session
Omar Chehab
No class (study day)
—
Final exam
Omar Chehab

Quizzes take place during the last 20–30 minutes of class. Quizzes and exams are in person; collaboration is not allowed.

Description

About the course. This course explores the linear algebra behind machine learning, including least squares, regularization, singular value decomposition, and optimization, with applications to regression, classification, clustering, and neural networks. Prior coursework in mathematical foundations and experience with numerical computing (such as Python) are required.

Logistics

Instructors
Grigorios Chrysos and Omar Chehab
Term
Fall 2026
Time
Tuesdays and Thursdays, 4–5:15 p.m.
Location
1800 Engineering Hall
TAs
Guruprasad Viswanathan Ramesh
Daiwei Chen
Karan Vikyath Veeranna Rupashree
Office hours
Omar — Nov 3 & 10, 5–6 p.m., 2540 Engineering Hall; Nov 17 & 24, Dec 1 & 8, 5:15–6:15 p.m., 1800 Engineering Hall.
Guruprasad — Tue, 10:30 a.m.–12:30 p.m., 307 Wendt Commons.
Daiwei — Thu, 9–11 a.m., 307 Wendt Commons.
Karan — Wed, 3–5 p.m., 307 Wendt Commons.