Machine learning explores the study and construction of algorithms that can learn from historical data and make inferences about future outcomes. This study is a marriage of algorithms, computation, and statistics so this class will have healthy doses of each. The goals of this course are to provide a thorough grounding in the fundamental methodologies and algorithms of machine learning.
Prerequisites: Students entering the class should be comfortable with programming and should have a pre-existing working knowledge of linear algebra (MATH 308), vector calculus (MATH 126), probability and statistics (CSE 312/STAT390), and algorithms. For a brief refresher, we recommend that you consult the linear algebra and statistics/probability reference materials on the Textbooks page.
Grading: (tentative) Your grade will be based on 5 homework assignments: HW0 (4%), HW1 (9%), HW2 (9%), HW3 (9%), HW4 (9%). There will be one midterm worth 20% and a final worth 40%.
In the past, CSE 446 was the undergraduate machine learning course, and CSE 546 was the graduate version. Over the years these courses have gotten closer and many undergraduates have opted to take the graduate version for a more challenging course. At the graduate level, some graduate students have sought a less demanding course to focus on research. To address the needs of our students, we are now offering two different versions of the course concurrently. A detailed overview of the difference between the courses and eligibility is below.
| Course | Section | Homework | Grading |
|---|---|---|---|
| 446 | Attend the section you are registered for. | A problems only. No credit will be rewarded for completing B problems. | You will be graded/curved against your peers in 446 only (on a 4.0 scale). Any attempt of the B problems will not influence your grade in any way. |
| 546 | Optional; you can attend any section of your choice. | A and B problems. | You will be graded/curved against your peers in 546 only. You will be expected to complete both A and B problems. Midterm and final exams will be the same as 446, but you will be graded against 546 students. |
Add codes are given out according to a centralized process organized by CSE. You can reach out to ugrad-adviser@cs.washington.edu or check out the following resources:
Unfortunately, we do not have the staff resources to support alternate exams, so attendance is mandatory in person for our pre-posted exam dates/times. For extenuating circumstances that prevent you from taking the midterm, e.g., a medical emergency with a doctor's note, we will allow you to use the final exam's grade to replace your midterm grade. However, this will increase the variance of your overall grade because it would rely heavily on your final exam grade.
Unfortunately, we do not have the staff resources to support alternate exams, so attendance is mandatory in person for our pre-posted exam dates/times. What we can do is allow you to take the final exam for next quarter's offering of 446/546. If you choose this path, you would complete our midterm and all of our HW assignments, and then we would temporarily give you an incomplete until we can process your final exam grade in winter quarter. If you'd like to take next quarter's exam, let us know before Dec 1.
It will have no effect on your grade. However, attempting the B problems can still be helpful for your learning.
Yes! Please contact the CSE advisors.
No. You will be evaluated based on whatever course you are officially registered for.
Yes, you are welcome to attend any section of your choice. However, please give priority in terms of space and time to those students registered for the section.
Yes. Your grade will be calculated based on both the A and B problems.
No. Curving each course will be based on an affine transformation of scores up to the discretion of the instructors alone and will not be publicly posted.