The prof was super nice and wanted us to learn. The class is in a flipped format so we had to watch lecture videos before actual lecture and during in-person class he would go over the material in more depth or have activities for us to do. While it was kind of annoying, the lecture videos were really good to have to review for the exams and I thought they were very helpful.

There are 7 homework assignments and honestly they were super hard for me. I definitely recommend going to office hours and his homework help lectures because otherwise I wouldn't have been able to complete them. The homeworks are genuinely pretty cool though, like simulating a google search or making a super primitive language model, but it really sucked working through them because they were really hard.

His midterm was pretty fair I thought but it's pencil and paper and almost all free-response where you have to write the output of the code or a code snippet. His final was much easier, a lot of it was on machine learning concepts and there was barely any coding, and it was NOT cumulative!

During discussion sections you have a notebook to work through with your assigned group which is graded only on completion. There is also a group project with the same group where you perform machine learning on a data set.

There are also lecture quizzes after every recorded lecture on Canvas, these have 1-4 questions. Some of them are kind of tricky so it kinda sucked when I would fail them because I got the only question wrong lol

Overall this class was pretty difficult mostly because of the homeworks, but he didn't write them someone else did so I'm pretty sure for every professor you have for this class the homeworks would e hard. I really liked having Murray as a professor though because he was really nice and helpful and lectured well.

Grading scheme:
homework 30%
discussion 10%
quizzes 5%
project 10%
midterm 20%
final 20%