Mathematics for Machine Learning 🧠

EECS 245, Winter 2027 at the University of Michigan

Instructor: Suraj Rampure, rampure@umich.edu
Lectures: TuTh 12-1:30PM, 220 CHRYS
Labs: Wednesdays at various times

Are you a computer science, data science, or statistics major who needs to take a linear algebra course and is interested in machine learning? Take EECS 245 in Winter 2027!

Content

Visit the course website from the current Fall 2026 offering. It contains links to our interactive course notes, lecture recordings, homeworks, labs, and exams.

Click to see the official catalog description.Mathematical foundations of machine learning and artificial intelligence, focusing primarily on linear algebra, along with select topics from calculus and probability. Topics include matrices, vectors, projections, spans, least squares, and eigenvalue and eigenvector problems. Includes mathematical theory as well as application to concrete data sets in a modern programming language.

Credit

The course counts for:

  • The linear algebra requirement for:
    • Computer Science majors (CS-Eng and CS-LSA*).
    • Data Science majors (DS-Eng and DS-LSA).
    • Statistics majors.
    • Artificial Intelligence minors.
  • The linear algebra prerequisite for:
    • EECS 442, EECS 445, EECS 448, EECS 474, EECS 476, and CSE 576.
    • Any 400-level STATS or DATASCI course with a linear algebra prerequisite.

*CS-LSA students don’t have a linear algebra requirement, but this course fits in the linear algebra “bucket” for the second math course; see here for details.

Prerequisites

This course has a math prerequisite and a programming prerequisite.

  • EECS 203 or Math 116 (or Math 215, 216, 275, 285, or 295), i.e. some math course after Calculus 1.
  • Some 100-level programming class (e.g. EECS 183, ENGR 101, ROB 102, or similar).

Logistics

In Winter 2027:

  • Lectures are in-person and recorded, and attendance is required. An alternative will be provided only for students with a documented class conflict.
  • Labs are in-person and attendance is part of your grade. Enroll in both a lecture and a lab.
  • There are 2 midterm exams and a cumulative final exam, with an opportunity to “redeem” lower scores on the midterms by doing better on the final (see the Fall 2026 syllabus for an example).
  • There may be occasional quizzes administered in the College of Engineering Electronic Testing Facility.

The full Winter 2027 course website, including the detailed syllabus, schedule, assignments, and resources, will be posted closer to the start of the semester.

Workload

This course has 11 weekly homeworks, weekly in-person labs, 2 non-cumulative midterm exams, and a cumulative final exam, which allows you to get back some of the points you lost on the midterms.

We asked Winter 2026 students how much time they spent on a typical homework. The most common response was 6–9 hours.

Winter 2026 self-reported homework time n = 50 responses
<3 2
3–6 9
6–9 29
9–12 9
12–15 0
15–18 1
Hours per homework

Our target is that the course takes a similar amount of time per week as EECS 203. Here’s what this semester’s students said about the workload relative to EECS 203:

Fall 2026 workload compared with EECS 203 n = 59 responses
1 1
2 15
3 24
4 17
5 2
Way less work than EECS 203 Way more work than EECS 203

Grade Distribution

See Atlas for the grade distribution from Winter 2026 and Spring 2026.

Course Evaluations

Questions?

If you have any questions about the course, don’t hesitate to contact Suraj Rampure at rampure@umich.edu. We’re happy to meet with students 1-on-1 to answer any questions they may have about the course.