STATS 102B
Introduction to Computation and Optimization for Statistics
Description: Lecture, three hours; discussion, one hour. Requisites: courses 100B (or Mathematics 170S), 102A, Mathematics 33A. Introduction to computational methods and optimization useful for statisticians. Use of computer programming to solve statistical problems. Topics include vector/matrix computation, multivariate normal distribution, principal component analysis, clustering analysis, gradient-based optimization, EM algorithm for missing data, and dynamic programming. P/NP or letter grading.
Units: 4.0
Units: 4.0
1 of 1
Most Helpful Review
Winter 2018 - Easy class bc you get tested on useless things. Literally don't understand why people love Miles, I've had minimal interactions with him but when I have interacted with him, he's pretty rude and sassy. He makes you feel small. Yes, I understand that he wants to spend some lecture giving lectures on life and shiz but also I WANT TO LEARN SOME USEFUL STATS MATERIAL. class is a joke
Winter 2018 - Easy class bc you get tested on useless things. Literally don't understand why people love Miles, I've had minimal interactions with him but when I have interacted with him, he's pretty rude and sassy. He makes you feel small. Yes, I understand that he wants to spend some lecture giving lectures on life and shiz but also I WANT TO LEARN SOME USEFUL STATS MATERIAL. class is a joke
AD
Most Helpful Review
Spring 2018 - Attention: Please read. Juana Sanchez is single-handedly the worst professor I've ever had at UCLA. Of course after 20 years of teaching, the rumors about how awful she is would make one initially hesitant to enroll in the class. However, she was the only professor Spring 2018, and I had to endure it with her. To give you some context, there were over 100 kids enrolled week 1, and during the final there were only 53 kids. OVER 50 KIDS DROPPED THIS COURSE. If this doesn't scare you, please keep reading. During the first lecture on week 1, we were given a pop quiz that counted for a grade. On her syllabus, she makes "participation" worth 8% , and the final worth 35%. If you chose, you can opt out of the in-class participation exercises and make your final 43%. I'm telling you now, this is a TRAP. And it was the first of many I experienced over 10 weeks. Do not drop the participation to make your final worth more, because the final is so anxiety-inducing. To give you an idea of how awful the homework assignments were, she would make you copy-and-paste code from R into a CCLE portal (instead of a rendered HTML or PDF like a normal professor). But if your code had ANY errors at all, she gives you a 0. In terms of the exams, Sanchez puts WAY TOO MANY QUESTIONS for anyone to finish in 50 minutes. Her testing policies are neurotic, and yes, she will read every word on your cheat sheet to make sure it follows ALL of her arbitrary rules. The midterm exams (each worth 23.5% of your grade) are a mix of MC, written response, and analyzing code. She will ask you to read through a page of code and point out ALL of the errors in the code AND fix them. (How was anyone supposed to do this??) Also, proofs are not off limits, and she won't hesitate to ask you for a proof on the exam! (Eyeroll) When it comes to exams, she also gives you the most inconvenient matricies to multiply. We had to multiply a (4x4) and a (4x1) BY HAND. To make it worse, each value had three decimal places. Perhaps the most annoying aspect of the exam is that she is very stingy with partial credit. Each exam question is only ONE POINT. So each exam is /15 points. (If you get two MC question wrong, you're already at a B, and it's worth 23.5%). To put things into further perspective, a 95% is a solid A. 94 or below is an A- . I bet you're wondering how the final exam could possibly be worse. Well, to answer that question: IT WAS ENTIRELY CODING. 100% of the final exam was done in R in a computer lab where we had to code algorithms to data sets she gave us. If you are not a strong programmer- this class is not for you. The final was open note (the only redeeming quality of this class), but the exam averages were awful throughout the quarter. The exam averages were: Exam 1) 57%, Exam 2) 61% , Final) 61% . Personally, I did above avg on one exam, slightly below avg on the other, and scored pretty well on the final, but somehow ended with a B- and do not think I deserved it at all. The smartest friends I had in that class also ended with a B- and I do not know anyone who earned a higher grade. 25/53 kids failed the final exam, so there is a high chance that ~50% of this class received a C+ or lower (just a rough estimate given my personal exam scores). If you are still somehow not convinced she is the WORST professor on this campus, keep reading buddy! On top of these awful exams and homeworks, she also is EXTREMELY patronizing and rude during lecture. She walks around the room while she lectures and actually kicked backpacks to walk through the aisles. I'm not exaggerating when I say almost every set of slides / worksheets she gave us had errors in them. Her lecturing style is very unorganized also. She scribbles all over the board illegibly, and does not post her handwritten notes online. There were SEVERAL instances where we would get emails from her saying "found an error in the notes, please correct it" and sometimes she wouldn't even tell us what the error was. She also frequently keeps you over the 50 minute lecture time. The only reason I got through this class was because the TA was decent. For someone who has been teaching for 20 years at UCLA, I find this all absolutely unacceptable. There is no amount of psychological counseling that can help one cope with the bullshit she puts you through for ten weeks. Absolutely avoid this class at all costs. If you are thinking about taking any other class with her, keep this in mind. She truly does not give a FUCK about any of her students- even the ones who show up nearly in tears to office hours. So glad I'm done with this class. See you in hell, Juana.
Spring 2018 - Attention: Please read. Juana Sanchez is single-handedly the worst professor I've ever had at UCLA. Of course after 20 years of teaching, the rumors about how awful she is would make one initially hesitant to enroll in the class. However, she was the only professor Spring 2018, and I had to endure it with her. To give you some context, there were over 100 kids enrolled week 1, and during the final there were only 53 kids. OVER 50 KIDS DROPPED THIS COURSE. If this doesn't scare you, please keep reading. During the first lecture on week 1, we were given a pop quiz that counted for a grade. On her syllabus, she makes "participation" worth 8% , and the final worth 35%. If you chose, you can opt out of the in-class participation exercises and make your final 43%. I'm telling you now, this is a TRAP. And it was the first of many I experienced over 10 weeks. Do not drop the participation to make your final worth more, because the final is so anxiety-inducing. To give you an idea of how awful the homework assignments were, she would make you copy-and-paste code from R into a CCLE portal (instead of a rendered HTML or PDF like a normal professor). But if your code had ANY errors at all, she gives you a 0. In terms of the exams, Sanchez puts WAY TOO MANY QUESTIONS for anyone to finish in 50 minutes. Her testing policies are neurotic, and yes, she will read every word on your cheat sheet to make sure it follows ALL of her arbitrary rules. The midterm exams (each worth 23.5% of your grade) are a mix of MC, written response, and analyzing code. She will ask you to read through a page of code and point out ALL of the errors in the code AND fix them. (How was anyone supposed to do this??) Also, proofs are not off limits, and she won't hesitate to ask you for a proof on the exam! (Eyeroll) When it comes to exams, she also gives you the most inconvenient matricies to multiply. We had to multiply a (4x4) and a (4x1) BY HAND. To make it worse, each value had three decimal places. Perhaps the most annoying aspect of the exam is that she is very stingy with partial credit. Each exam question is only ONE POINT. So each exam is /15 points. (If you get two MC question wrong, you're already at a B, and it's worth 23.5%). To put things into further perspective, a 95% is a solid A. 94 or below is an A- . I bet you're wondering how the final exam could possibly be worse. Well, to answer that question: IT WAS ENTIRELY CODING. 100% of the final exam was done in R in a computer lab where we had to code algorithms to data sets she gave us. If you are not a strong programmer- this class is not for you. The final was open note (the only redeeming quality of this class), but the exam averages were awful throughout the quarter. The exam averages were: Exam 1) 57%, Exam 2) 61% , Final) 61% . Personally, I did above avg on one exam, slightly below avg on the other, and scored pretty well on the final, but somehow ended with a B- and do not think I deserved it at all. The smartest friends I had in that class also ended with a B- and I do not know anyone who earned a higher grade. 25/53 kids failed the final exam, so there is a high chance that ~50% of this class received a C+ or lower (just a rough estimate given my personal exam scores). If you are still somehow not convinced she is the WORST professor on this campus, keep reading buddy! On top of these awful exams and homeworks, she also is EXTREMELY patronizing and rude during lecture. She walks around the room while she lectures and actually kicked backpacks to walk through the aisles. I'm not exaggerating when I say almost every set of slides / worksheets she gave us had errors in them. Her lecturing style is very unorganized also. She scribbles all over the board illegibly, and does not post her handwritten notes online. There were SEVERAL instances where we would get emails from her saying "found an error in the notes, please correct it" and sometimes she wouldn't even tell us what the error was. She also frequently keeps you over the 50 minute lecture time. The only reason I got through this class was because the TA was decent. For someone who has been teaching for 20 years at UCLA, I find this all absolutely unacceptable. There is no amount of psychological counseling that can help one cope with the bullshit she puts you through for ten weeks. Absolutely avoid this class at all costs. If you are thinking about taking any other class with her, keep this in mind. She truly does not give a FUCK about any of her students- even the ones who show up nearly in tears to office hours. So glad I'm done with this class. See you in hell, Juana.
Most Helpful Review
Spring 2024 - Great class. It was his first time teaching 102B, but he was able to teach all of the topics and explain them really well. The material is almost identical to Math 156 and its a cool and relevant class for understanding machine learning. Lectures are recorded and aren't mandatory. HWs really help you understand the concepts and they're easy enough to do in one sitting. Exams are also not so bad, and he emphasizes the importance of mental health and not worrying too much about your grade with lots of motivational speeches. Michael and Miles are my top 2 for stats profs, so I was pleased to take his class.
Spring 2024 - Great class. It was his first time teaching 102B, but he was able to teach all of the topics and explain them really well. The material is almost identical to Math 156 and its a cool and relevant class for understanding machine learning. Lectures are recorded and aren't mandatory. HWs really help you understand the concepts and they're easy enough to do in one sitting. Exams are also not so bad, and he emphasizes the importance of mental health and not worrying too much about your grade with lots of motivational speeches. Michael and Miles are my top 2 for stats profs, so I was pleased to take his class.
Most Helpful Review
Winter 2017 - Zhou is one of my favorite professors in the Stats department and I like the way he teaches math. Usually professors who teach math write messily on the board and doesn't pay attention to whether students understand the lecture. But Zhou was very clear throughout every class and always made sure to review the pre-req concepts. He doesn't shy away from giving proofs but you always learn a lot about the details of how the math works, which is not emphasized in many other Stats classes (ie. 101 series). Coding isn't as important in 102B as it is in 102A or 102C, since this class is mostly about linear algebra and optimization. Homework problems (this year's TA gave away the solutions) and exams were focused on concepts rather than computation, but they weren't nearly as challenging as Christou's. I would recommend Zhou because he was a good lecturer and knew his stuff inside out.
Winter 2017 - Zhou is one of my favorite professors in the Stats department and I like the way he teaches math. Usually professors who teach math write messily on the board and doesn't pay attention to whether students understand the lecture. But Zhou was very clear throughout every class and always made sure to review the pre-req concepts. He doesn't shy away from giving proofs but you always learn a lot about the details of how the math works, which is not emphasized in many other Stats classes (ie. 101 series). Coding isn't as important in 102B as it is in 102A or 102C, since this class is mostly about linear algebra and optimization. Homework problems (this year's TA gave away the solutions) and exams were focused on concepts rather than computation, but they weren't nearly as challenging as Christou's. I would recommend Zhou because he was a good lecturer and knew his stuff inside out.
AD
1 of 1
