STAT W3107

weekly psets: ended up with 10 total, lowest grade dropped midterm: no review, 1 sheet of formulas, 6 problems. I think one person finished the whole thing. final: not cumulative, review session, 1 sheet of formulas, choose 5 out of 7 or 8 questions. This was more doable than the midterm, but still challenging.

Problem Sets (20% of grade) - taken from the book, and usually given every two weeks. There were 7 problem sets for the semester. Midterms (20% each) - can be challenging, but no real big surprises. Final (40%) - it was okay. There is a curve applied to the overall grade.

Problem sets every week (difficulty ranging from easy to medium, about 10 problems a week), one midterm and one final (easy with an equation sheet). Comparatively lax relative to other stats courses.

Simply put, this class was hard. I took it as part of the probstat requirement for Information Science, and if I had the choice, I would definitely go back and take 4150 or 3600 instead. You need to come in with a solid foundation in probability, and if you don't have that, you're kind of screwed from the start. There is a quick review of probability in the first few lectures. The material is challenging, and Prof. Elbarmi does his best to teach it, but sometimes it just isn't enough. The lectures are extremely organized. They were dry, but the material is to blame, not the professor. He systematically goes through definitions, theorems, and examples. Many of these are straight from the book. If you get to class early, you can ask him questions, and he'll answer them on the board. The homeworks, aside from those in the beginning for the probability review, were always 5 questions from the textbook. Five questions sounds short but these psets were the bane of my existence for the entire semester because they were just too damn hard for me. They took me hours, but I am sure they were easier for other people who were more comfortable with the material. The exams were challenging, but the material did not stray far from lecture examples and homework questions. The professor was pretty unresponsive to emails. The TAs were only slightly better. Just go to office hours or ask the pro…

A rigorous introduction to statistics. The presentation of material was straightforward, with definitions, theorems, and occasional proofs given in class and in the notes. Some interesting examples in the homework and in class. Covered probability, estimators, and hypothesis testing, focussing on theory more than application. Not too difficult if you have a solid foundation in math and are comfortable with proofs. Professor Mukherjee is a good lecturer in that he is straightforward and thorough.

Neath is a satisfactory instructor. His classes are standard lectures in which he writes notes on the board and occasionally asks questions directed to the class. I found the notes to be quite helpful sometimes. The textbook (Hogg, Craig, McKean) was rather dry and his notes helped to clarify the important parts and provided examples. His explanations are also generally helpful. Finally, he is generally responsive to emails in a timely manner. Overall, not an amazing professor, but he's pretty good. He also smiles a lot and wears jeans all the time.

Libor, as one of the few Columbia quantitative professors who has a working understanding of the English language, shoots pretty much par for the course. He’s generic: wonderfully concise yet often painfully by-the-book. He’ll do a fantastic job teaching the material; you’ll be studying mostly from your notes because the book is two exits beyond awful. This being said, don’t get your hopes up for an engaging student-teacher interaction between you and Libor. He is a Statistics professor after all. Overall, Libor is a good professor who you should not shy away from.

about 10 problem sets(5 questions each); 1 midterm (1 page front and back of notes allowed); 1 final (2 pages front and back of notes allowed). If you attend lecture, not bad at all.

Hammou is one of the best professors I have had at Columbia thus far. He truly cares about his students and took the time to actually learn our names in a large class of 50+ students. His lectures/notes are organized and beautifully presented and he encourages class participation during the lecture. He makes sure to ask if there are any questions quite frequently throughout the lecture. I highly recommend him; he makes learning statistics a joy!

This is one of the best professors that I had at Columbia. Everyone complaining about the bad grades are either lazy or stupid. First, on the complaints about no review for midterm/final: The examples he gives in lecture are practically identical to the homework, which are in turn practically identical to the midterm/final (and in fact there have been exact copies of hw problems on these tests). He lets you bring in FOUR sheets of notes for the final and I think at least two for the midterm. They can be typed and they can contain worked problems. Put the examples from class on them and the very few formulas you need. On grades: He gives Cs because people (myself included) get close to 100%s in the class. If several people have a 98 in the class and you have a 65, you deserve a C. You are not entitled to a good grade in a class where you do not understand the material and, for this class, the fault is on you, not the professor. On his teaching style: He prepares and presents beautifully organized notes in class and these are all you need to do well. The book is terrible and confusing and I rarely read it. If you skip class and try to learn out of the book you are putting yourself at a terrible disadvantage. He also encourages participation in class, which keeps you awake and helps you to make connections yourself. He is also very nice and learns most of the names of the student…