Covers probabilistic, algebraic, and geometric data models, mathematical analysis of learning algorithms and optimization methods, and ML applications.
Past student reviews · Grigoris Chrysos
Homeworks and exams were challenging but commensurate with the proof-based, graduate-level nature of the class.
Derive and apply mathematical tools for machine learning from probability, statistics, linear algebra, and optimization
Perform mathematical analysis and characterization of generative and discriminative models
Perform mathematical analysis of machine learning algorithms
Perform derivation of basic machine learning error bounds and related performance analysis
Read and understand theoretical papers from machine learning conferences
Source: UW Guide 1· UW Guide 2
All recorded terms · compare terms & instructors
Graduate/professional standing
“Used by” includes alternatives; linked courses may have other requirements. This is a best-effort interpretation; check the catalog requirements above.
Raw average: 1.21/5 from 29 quality ratings. The adjusted rating blends this with the UW review average (3.66/5), weighted as 20 additional ratings. Smaller samples stay closer to that average. Each captured review is counted once in the prior; this does not correct who chooses to leave a review.
For this course: 2.5/5 raw quality · 4.5/5 difficulty · 2 reviews
RMP profile ↗ · All captured review dates; profile matched by name.
Past student reviews · Grigoris Chrysos
Grigoris Chrysos taught a proof-based course on ML foundations covering tensor decompositions and statistical learning theory. Students reported conflicting experiences: one described clear lectures and challenging assessments, while another criticized illegible notes, unrelated homework, and the professor's absence for a month.
Recent recorded grades — Spring 2025: 3.54 GPA, 69.2% A/AB (n=39 letter grades).
Past student reviews · Ramya Vinayak
Ramya Vinayak required strong backgrounds in probability, statistics, and optimization, making the course challenging for students lacking these prerequisites. She explained complex concepts like uniform convergence and Rademacher complexity without hand-waving, tying proofs back to actual learning algorithms. While the workload was heavy, students noted they learned valuable concepts from the rigorous instruction.
GRIGORIS CHRYSOS is recorded teaching in Spring 2025. Recorded history may be incomplete and does not establish a future schedule.
RAMYA KORLAKAI VINAYAK is recorded teaching in Spring 2023, Spring 2024, Spring 2026. Recorded history may be incomplete and does not establish a future schedule.
ROBERT NOWAK is recorded teaching in Spring 2016, Fall 2017, Fall 2018, Spring 2020, Spring 2022, Spring 2023. Recorded history may be incomplete and does not establish a future schedule.
Recorded history may be incomplete and does not establish a future schedule.
Schedule loads here as you scroll.
| Section | Mode | Enrolled / capacity | Waitlist |
|---|---|---|---|
| LEC 001 | Classroom Instruction | 0 / 45 | 0 |
| LEC 001 | Classroom Instruction | 0 / 45 | 0 |
Times are Central. Select a meeting for details; export includes recorded dates for the selected sections. Enrollment reflects scan time.
Past student reviews · Grigoris Chrysos
This graduate-level course focused on the mathematical foundations of machine learning, covering topics like tensor decompositions and statistical learning theory with clear lectures.
Recent recorded grades — Spring 2024: 3.74 GPA, 87.0% A/AB (n=23 letter grades); Spring 2025: 3.54 GPA, 69.2% A/AB (n=39 letter grades); Spring 2026: 4.00 GPA, 100.0% A/AB (n=8 letter grades).
Past student reviews · Grigoris Chrysos
Homeworks and exams were challenging but commensurate with the proof-based, graduate-level nature of the class.
Past student reviews · Grigoris Chrysos
One student found lectures difficult to follow due to the instructor's handwriting, noted that homework was unrelated to class, and reported the instructor disappeared for the whole last month.
Latest available · Spring 2026— not enough history to project Spring 2027.
Through Spring 2026
Not enough comparable courses for Spring 2027 in UW–Madison.
Descriptions, prerequisites, and recorded course offerings.
Observations at scan time; dates do not imply when a catalog change took effect.
Mathematical Foundations of Machine Learning
Recorded 2026-10-14Mathematical Foundations of Machine Learning
Recorded 2026-10-14[
{
"run_id": "20261014T203208-0851c9a1",
"semester": "1274",
"observed_at": "2026-10-14 17:50:09.728220+00:00",
"offering_id": "1274:266:023639",
"course_id": "COMPSCI/ECE 761",
"course_uid": "course_a7ce5ee394c07ac642067916",
"term_id": "1274",
"source_course_id": "023639",
"source_subject_id": "266",
"title": "Mathematical Foundations of Machine Learning",
"credits_min": 3,
"credits_max": 3,
"typically_offered": "Occasionally"
},
{
"run_id": "20261014T203208-0851c9a1",
"semester": "1274",
"observed_at": "2026-10-14 17:50:09.728220+00:00",
"offering_id": "1274:320:023639",
"course_id": "COMPSCI/ECE 761",
"course_uid": "course_a7ce5ee394c07ac642067916",
"term_id": "1274",
"source_course_id": "023639",
"source_subject_id": "320",
"title": "Mathematical Foundations of Machine Learning",
"credits_min": 3,
"credits_max": 3,
"typically_offered": "Occasionally"
}
]Original comments behind the course and instructor summaries.
Recorded grade distributions by term, section, and instructor.
Explore recorded gradesRecorded 2026-10-14
Recorded model configuration, reasoning, and tool conversations.
{
"model": "Qwen/Qwen3.8-27B",
"model_revision": "1d4bf0f2ff6012fd82039f2fa52739d0dd7c60c0",
"task_version": "17",
"output_id": "a5e5ca6b4385f190e4624a5fb133c7341d9b6c9156ef3bd03781b70e400f0b9b",
"requirements_status": "valid",
"dataset_revision": "fc1811fa52e581ea57a35c0cf3343f8740a8ffbf",
"observed_at": "2026-10-14 17:50:09.728220+00:00"
}