This class is a pretty useless class, since there is a huge lack of practice material. It doesn't help that the homeworks barely test on what is covered in class, leading to like 5 people attending live lecture everyday, while there are 100 people in the class.

This class is focused on a bunch of ML models, like M146, except there are no derivations, and you're basically just using models from sklearn / other ML libraries, and running them to see results. To me, this seemed extremely stupid, since I had no idea what any of the models were doing inside. Also, we never even learned how to do random search or grid search for hyperparameter tuning, which made this class even stupider.

Honestly, this class material could be learned in like 1 week if there was a good textbook and syllabus to follow, since we barely covered anything in depth in the homework or the tests. There were only 3 homeworks and 3 projects, each probably took less than 2 hours, so a very light workload.

The exam was the worst part of this class. If we actually knew anything, the exam would be easy, but because of no practice problems in lecture, like 1 practice problem in discussion, and no textbook, it was impossible to practice for the exam. If there was 1 practice exam, I would have understood what I was weak on .. but no, hence the exam was hard even though I could have studied all the relevant practice problems in like 30 minutes.

Also, Piazza communication is super weak here, questions were left unanswered for weeks and hastily answered before the final. Not a good look.

All in all, a class not worth taking. If you want to learn how to implement ML models, spend like 5 minutes on sklearn. If you want to learn the inner workings of basic ML models, take M146 (you def do NOT learn it in this class). If you want to learn the inner workings of neural networks, take ECE 247, or spend 15 minutes watching a 3 blue 1 brown video. It's not even an easy A since the test at the end is a total crapshoot and worth 40% of your grade; if you want an ez class take CM122. Rant over!

Prof. Mirzasoleman is a very nice and a great professor. She is always calm. I really enjoyed her lectures. She was also always available to help students and answer their questions. She is clearly an expert in this area and she enjoys explaining them to students. The exam was hard but it is a fair one if you attend the lectures and do the homework. I wish there were more examples in the lectures or discussion sections. But, overall, I enjoyed the course and I highly recommend it. It gives a very good high-level picture of ML and Data Science.

The course covers A LOT of topics but I liked it since it gives a good overview of different topics in ML and Data Science without going to math and details. Prof. Mirzasoleiman is also a great professor and she clearly cares a lot about her students. She also explains the topics very well and engaging. There was only one exam which was a bit hard but they were fair and generous in grading. I wish the TAs were more active in discussion section (maybe solving more examples?) and more responsive in Piazza. Overall, I highly recommend this course. In particular if you don't like math but you want to get a high-level picture of data science, this is a great course.

The material is pretty interesting, it's like the more applied/less theoretical version of M146 (machine learning). The class was pretty disorganized and the grading wasn't great, but I guess that's to be expected of new classes, it'll probably be better in the future. My TA Lionel was awesome, really cares about the students and puts a lot of effort into making discussion good. I'd recommend.