ECE-689
ADV TOPICS IN DEEP LEARNING
Not in Fall 2026
Term
Overview
Feedback is mostly positive. The strongest signal is that students generally rate the course well. Difficulty runs on the high side even without a single dominant complaint theme. Best for students who can handle a demanding pace without needing constant hand-holding.
DepartmentELEC&CMP
Terms offeredSpring
Typical enrollment38–52
Semesters of data2
6.9
Hrs / week
46
Responses
90
Enrollment
51%
Response Rate
Evaluation Scores
Overall quality
Teaching, content, and experience combined.
4.5
Intellectually stimulating
Challenges students to think deeply.
4.5
Instructor effectiveness
Explains concepts and facilitates learning.
4.7
Difficulty
Higher means harder.
3.7
Feedback Analysis
Feedback Analysismedium
Analysis based on student evaluations
Based on 45 comments across 2 sections
Feedback is mostly positive. The strongest signal is that students generally rate the course well. Difficulty runs on the high side even without a single dominant complaint theme. Best for students who can handle a demanding pace without needing constant hand-holding.
Student Reports
How hard is the A?
A is doable but not automatic
The signal here is more do-the-work-and-you-should-be-fine than easy-A chatter. Students do not describe the A as automatic, but the evidence also does not paint grading as punishing.
Homework Load
Heavy homework load
Homework load is one of the clearest friction points. Students repeatedly describe assignments, readings, or problem sets as time-consuming.
Lecture Load
Lighter lecture burden
Student comments describe this as more discussion-, seminar-, or workshop-driven than lecture-dependent. The lecture burden itself does not sound like the main source of friction.
Strengths
• Instructor ratings are strong even when the comments do not cluster around one obvious positive theme.
Tradeoffs
• Difficulty runs high even when comments do not settle on one dominant complaint.
Best fit for
Best for students who can handle a demanding pace without needing constant hand-holding.
Watch out for
• A large share of the evidence comes from one instructor's version of the course, so this may not generalize cleanly.
Student Responses
Latest trend in deep learning
Spring 2025 · Tarokh, Vahid
A survey in Deep Learning. Learned about new physics based methods, sampling methods, and current state of the art generative models.
Spring 2025 · Tarokh, Vahid
In this course, I developed a solid foundation in deep learning by learning key models and theoretical concepts, such as neural networks, convolutional networks, and optimization methods. I also gained practical experience applying these models to real-world datasets through homework assignments, which helped me understand how to preprocess data, tune models, and evaluate results. Finally, I improved my ability to think critically about model design and performance, allowing me to approach machine learning problems more analytically.
Spring 2025 · Tarokh, Vahid
Researching on my own and finding out about recent breakthroughs.
Spring 2025 · Tarokh, Vahid
Some advanced knowledge in several deep learning topics.
Spring 2025 · Tarokh, Vahid
Rating History
Rating history
Error bars show \u00B11 std dev
| Term | Instructor | Overall | Difficulty | Hrs/wk | Enrolled |
|---|---|---|---|---|---|
| Spring 2025 | Tarokh, Vahid 1.5Rate My ProfessorsQuality1.5Difficulty4.5Would retake0%Based on 2 ratingsClick to view on RMP → | 4.5 | 3.7 | 6.8 | 52 |
| Spring 2024 | Tarokh, Vahid 1.5Rate My ProfessorsQuality1.5Difficulty4.5Would retake0%Based on 2 ratingsClick to view on RMP → | — | — | 6.9 | 38 |
Instructor
Tarokh, VahidELEC&CMP
Also teaches
ECE-685D INTRO TO DEEP LEARNING4.0