I took this class during the pandemic and everything was taught online so my experience might be different than a regular semester. Having said that, she is probably the worst CS professor I have ever got.

First of all, her class is basically a math class rather than an ML class. I get that math is important and is the core of ML, but it's not the ONLY thing. All her lectures are just her writing math (in horrible handwriting that you probably won't understand) and talking about math. She doesn't care about explaining the algorithm or give the intuition behind why and where we would use the particular methods. She just cares about putting down the math and proving shit that you probably don't need. I literally had to watch Andrew Ng's lecture on SVM that he taught in a GRADUATE class at Stanford to understand what the hell the professor was talking about. The math would probably not be daunting if she just knew how to teach. If you want to pass her class then you need to be a math genius and follow another ML course alongside to get an A in the class.

Also, homework is basically the same thing. It's all about math and proving this proving that. There's a coding portion to the homework but those questions test your coding abilities rather than any ML knowledge. Most of the homeworks contained questions where you had to use a library that is practically impossible to figure out how to use. And the worst part is nor the professor or the TAs talk about it.

However, her tests were doable and we were given 24 hours to finish them. So in the end I managed to get an A only because of online schooling and take-home exams.

Prof. Dolecek is a good person, and she is very knowledgeable when it comes to the course material. I want to make this clear that she is NOT a bad person or mean or anything.

That being said, there are some points that you should know if you were to choose her lecture (especially for remote learning):

1. She didn't use zoom. All lectures are pre-recorded and posted on CCLE for my quarter.

2. She has terrible, terrible handwriting. Sometimes you cannot tell subtractions apart from multiplications (she writes · and - really casually), also from time to time her writing becomes unreadable and you have to rely fully on listening.

3. For some reason, in the middle of the quarter she switched from ball-point to highlighter to write on her slides, just when you think her handwriting cannot get any worse...

So pretty much her handwriting has made this course harder than it should be, and the highlighter is plain suffer for remote learning. But again, Prof. Dolecek is a good person, she would answer questions and can explain stuff for you when you are stuck.

Overall this is a math class, with minimal programming questions sprinkled in. Midterms, final, lectures, and 75% of the hw questions were all math proofs/questions. However, the exams were pretty easy and straightforward. They were open book, which was a lifesaver since the questions were heavily based on her lectures. Recommend taking 10/10

This class quite challenging.Be prepared for a lot of math (matrix calculus, probability, convex optimization). Professor Dolecek goes through all the mathematical proofs in detail. But you still need to spend some extra time to learn and understand them. (Also her handwriting makes her slides unreadable, so do take notes during lectures otherwise you won't understand what she was writing).
The homework is usually 5-6 problems, with 1 or 2 asking you to implement an algorithm you learned in class with python or matlab. It is definitely time consuming, especially the programming problems. I personally spent ~10hrs per week for the homework (partially because I'm not familiar with python and matlab). However, after struggling through all the problems, I did learn a lot.
The midterm was a lot easier than the homework. So don't get too stressed for that. (I can't speak for the final though bc I opt-out of it).
Overall I'll recommend this course. It's challenging but you learn a lot.