OK this review may destroy all the good impressions you got from the other reviews, but I have to tell u I am telling ONE HUNDRED PERCENT TRUTH!!!!!

Dr. Sarrafzadeh clearly knows the contents of the course very well because I have to admit his lecture contents are acceptable. I enjoyed his lectures because he always tried his best to introduce the algorithms and lead us through the steps. As other reviews mentioned, he came with a bit of "sarcasm," and this really made his lectures intriguing. His homework questions were fair considering the time I spent working on them and the TAs' generous grading policy. There are usually 6 questions per problem set and I spent roughly 6h to finish each week. His midterm was comparatively easy, and I got almost 100 out of it. Until now u are seeing positive comments right? Get prepared here comes the bullshit part.

Note, I got almost 100% before I took the final exam. But for the final, he designed, like the midterm, 2 sets of problems, one for domestic students and one for international students (there were 2 exam time windows, one at 8am, one at 8pm). I took the 8pm test and got 70/100. There were 23 students (most of which from Asia) who took the 8pm final, and we had a group chat. I checked with most students in the group chat and found out 70/100 was actually a pretty decent grade because most of my peers got a score around 60-65.

Now you may think the average for the final should be around 65 and Majid would curve up the grade . Well you are wrong, or I have to say more than wrong. Later I received a B grade and he told us he would not curve because the final grade was "unusually high". As a result, I asked the Dean for help and got the distribution of the final grades. It turned out domestic time window got a mean of around 88, and international 78. I could not figure out why was the median 78 given that I had already spoken to most of my classmates taking the 8pm test and they received scores around 65??? He gave out 62/180ish A's and it turned out almost none of them were from the 8pm test. He insisted that the 2 tests had same difficulty level, and international students did poor on the exam simply because they did not master the materials well. But I got 100/100 before the final and I still screwed the final to such an extent. So I personally do not think he offered a fair set of tests.

"WHAT A GOOD ALGORITHM," quoting one of my classmates

Designed two sets of tests with one easier for domestic students and one harder for international students?
What a good strategy!

Boring class.

Took it with TA Sam Lee
definitely made it easier for me!
GOAT

Would not recommend. Majid is a spiteful person who made my blood curdle every time he spoke.

Maybe you want to skip this review because I typically don't do well in my CS courses, so my perspective might not be representative of the average CS major. However, I think this is the class that will finally make me drop out of the major for real.

I'm complaining a lot, but I really just feel defeated. Throughout the quarter, I went to and paid complete attention in every lecture and discussion, asking lots of questions and taking notes. The lectures aren't recorded, so if you don't go, you're basically screwed. Ironically, I feel like he explains things very well, and his lectures and the TA discussion slides are a lot easier to understand than the textbook. Also, sometimes he drops major hints for the midterm and final, but these are pretty rare.

The homework sometimes took me 6+ hours to finish, but it was pretty doable if you start early like I did and asked TAs/had a good study group. The graders sometimes took off points for random reasons, but I would generally say that they were pretty lenient.

My problem is the way the exams are graded. Some of the problems (mainly the creative ones) are looking for very specific keywords and steps in your algorithm, and if you don't have those, you basically get a 7/20 on that problem or worse. You are expected to give a proof and time complexity analysis for most problems but these are barely worth any points at all. The final for this class was the worst score I have ever gotten on any CS exam, and while the midterm was much more reasonable, the grading rubric was also pretty unclear. I just think it wasn't really fair at all because one of the questions was extremely vague and no one could tell what it was even referring to, and people got points off for doing what was essentially said in lecture. Many of the other questions were similarly vague, and the rubrics were as well. I nearly got a zero on 3 questions and most of them just say that my algorithm is incorrect without telling me why. I just feel like the questions on that exam went way beyond what was taught in lecture, and the official correct answer was something that no one could have come up with. In these cases regrade requests are your only hope.

I don't think this professor is even usually this bad. However, if you're like me and find CS difficult, you're probably better off taking it with a different professor or doing an alternative. The tests can be a complete RNG and the grading was...something. Based on my calculations I will probably get a low C in this class if I factor in the final, and I'm really worried. I don't want to discourage anyone who (unlike me) can probably handle this class fine, but I really feel like I did everything I could this time and got nothing in return.

I personally think Majid is a good professor. His lectures were clear and paced really well (not too slow, not too fast, not too boring, not too many jokes). The HWs were long and hard but doable if you put time into it (and they're graded very leniently). For exams, he gave points quite leniently as long as you wrote something for proofs/time complexity.

I feel like Majid really cares about the students. For the final, he actually changed the grading rubrics of two questions in favor of the majority of the students due to our feedback, which is rare to see. I also like how he gave a large percentage of A's because of the high avg/median in the exams (midterm and final were both around 90) instead of curving down the class.

This professor does not give a shit about the students at all. No matter what he says or did throughout the quarter, he does not care at all and you should NOT take it with him - AVOID AT ALL COSTS. He emphasizes understanding the content over memorization and that as long as you make sure to review his lectures, complete the projects, complete the homeworks, and study the sparse resources he gave us (1 practice exam that he didn't write), you will do well in this class.

This is a complete and utter lie. I aced the homeworks and projects alike as I understood all the content he went over in class, and then this horrible, uncaring professor goes ahead and on an exam worth 40% of our grade, puts questions that have us doing calculations he never even mentioned in class, says that the exam will be as long as or shorter than the practice exam in terms of number of questions, and then goes ahead and makes it much longer, and only gives us 90 minutes for the exam (which is worth the same as a final exam which is generally 3 hours and gives us half the time).

This professor hands down does not give shit about the students. If it wasn't for the TAs seeing how shitty the professor was and making sure to give out extra points, a lot of people would be leaving the class with Cs SOLELY because of the final. I have never had such a hypocritical and uncaring professor at UCLA and I feel disgusted. Even Eggert makes sure to at least curve the class at the end so even if you get screwed by the scores, you have a chance to end up with a better grade than you expected. I know some people will downvote this answer, and I can tell you it will likely be those students who took CSM146 as they were already taught the content on the questions that Majid never taught us and thus easily aced the final (Side note, it is because these students did well that Majid ended up not curving because in his inane and broken logic, its fair to put questions only those with previous experience in an ML course could answer).

In conclusion, this professor is truly terrible and avoid taking ANY class with him if possible (his CS 180 class was dogshit as well: he down curved the class after explicitly saying he wouldn't).

Majid was a good lecturer—spoke clearly, made the content approachable, and was entertaining.

Homework was around six questions per week and was graded primarily on completion. Homework solutions were online for almost every problem. Though, if you want to derive the solution on your own and fully understand each question, expect the homework to take a while to complete (on average >1 hour per problem). Homework questions seemed more difficult than exam questions, so if you understand the homework, you should do well on the exams.

With respect to the midterm and final, partial credit is key in this class: if you write the correct keyword in your algorithm or proof, you can get a majority of the points while still having a flawed algorithm. You may even have a completely incorrect algorithm, but if you explicitly cite paradigms taught in class, you will earn many points.

Overall, this was a good class, and I'm glad that Majid was the professor. I highly recommend diving into the material since this exact content is what needs to be mastered for leetcode/job interviews. Good luck!

You would think that teaching a CORE CS class would institute someone who gives more of a fuck about his class. He is almost unreachable, and when he is, he defers you to his TAs (who were also pretty iffy). The Homeworks take way too long (3-5 hours) if you wanna do them right and usually you do them wrong and they do not really pertain to his lecture material. He is a decent lecturer but will only take up 50-70% of the 2 hour time and he even removed a part of the syllabus (Bellman Ford Alg.) to make up for his laziness. The grading and exams are abysmal because the questions are worded horribly to convey his intended answers. Ive gone from 5/20 to 20/20 on a question because of a misinterpretation of an exam question, where the TAs later had to fix. This was also the case for the final. I could go on and on about how this guy should NOT be teaching this important of a class but you should read my fellow peers' reviews on him (esp. this shitshow of a quarter: F23)