589 Machine Learning Spring 2021

This is the public page for COMPSCI 589 Machine Learning, taught by Justin Domke in Spring 2021.

Announcements

📌 We really recommended that you read the ‣.

📌 Our final will be on May 12 9:00am-11:00am.

📌 If you are a graduate student and would like to receive a SAT/UNSAT grade for this course, please fill out the following form. You may select a conditional cutoff (e.g., SAT/UNSAT if and only if your grade is less than a B+). **Graduate students only, please. (**Undergraduates should select pass/fail grading through SPIRE.)

May 11: All scribe notes available.
May 10: Quiz 4 solutions are available:
May 10: Some students have asked for a solution to quiz 4. It isn't possible to give a single global solution, as there are several different versions of the quiz. However we will prepare one version of a solution tonight which will be posted tonight.
May 6: We have received many emails from students struggling to upload their .zip file for Assignment 6 due to the file being too large. Our current recommendation is to simply exclude the "faces" dataset folder from your .zip file (it is okay if your runme file does not work due to this, as long as we can figure out where the folder belongs).
May 4: Justin will have office hours on Thursday May 6 from 4pm-5pm and Monday May 10 from 4pm to 5pm. (If these sessions are crowded I will add more.)
May 4: Ashish will hold extra office hours today from 12:00-1:00 PM.
May 3: The HW 6 deadline has been extended to May 4th, 5:00 PM.
May 3: We made one small change for HW 6, question 5: You are no longer required to submit results for k=150 and k=300. If you already prepared your solutions including these values that is perfectly fine — no need to remove them.
Apr 30: You may use sklearn's PCA implementation for the final assignment. (Apologies for the confusion— we made this change shortly before releasing the assignment and neglected to remove all references to the prohibition on using PCA that I wrote before the semester began.)