Rachel Schutt

You're either going to love this course, or hate it with a fiery passion. I loved this course. I loved it because I learned R, Python, and SQL, as well as had the opportunity to speak with professionals from leading tech companies who came in as guest lecturers. This course should not have "Introduction" in the title though. Using that word is wildly deceiving. I'm currently in a technical program but come from a not-so-technical background with limited programming experience, and this class was hard as hell. I feel confident in saying that I spent between 40 - 60 hours every single week of the semester working on something for this class. To give you a general idea, the course started with 100 students (max) on the roster with ~50 students waitlisted, we ended the semester with ~60 students on the roster. I think that says something. In regards to prerequisites, there technically are none, but this is a direct quote from the syllabus: "If you want to take this course but have holes in your background, then it’s your responsibility to fill in those holes by collaborating with others or studying and reading." And I had holes. Lots of them. This course was supposed to be split between Rachel and Kayur, but once Rachel landed a job at News Corp she rarely came to class. I believe she had to travel a lot for News Corp. For example, she Skyped in for our midterm presentations from…

Rachel is a grad student in Statistics and a competent instructor. She went a little fast sometimes but covered all the material and gave useful examples. She evidently puts a lot into her teaching and cares about how her students do. A solid section to take if you have a choice.

This is without a doubt the worst class I have ever taken at Columbia. It was taught by two completely incompetent adjunct professors, Rachel Schutt and Kayur Patel. Rachel pretty much completely checked out pretty early in the semester once she got a better job at News Corp. She couldn't even be bothered to show up for our midterm or final presentations. I imagine Rupert Murdoch is paying her enough, so I encourage Columbia to fire her. As our main point of contact for this course with Rachel being constantly gone, Kayur failed spectacularly. He was incredibly rude to students who tried to reach out to him and consistently would email us an hour or two before a lecture with required reading. My favorite bit of classiness from Kayur came on Thanksgiving, when he emailed some of us to tell us we'd be giving our final presentations in less than a week without any prior notice. In fact, the only positive thing I have to say about Kayur is that unlike Rachel, at least he showed up. Put simply, the assignments for this course were horrible. They consisted of either mundane tasks nobody should ever have to do (sorting 100 numbers by hand was a highlight) or tasks that we had simply not been prepared to tackle. They never taught us much of anything, but certainly asked us to do quite a lot (for example, asking us to do a full implementation of Naive Bayes without ever even mentioning…

your best bet for stat. she's dedicated and actually teaches, which is rare in the stat dept. she grades fairly and theres a decent amount of work, but any section will have that. her problems at the beginning of each class help clarify what happened last class, and her homework is a bit lengthy, but doable. the stat requirement sucks for econ majors, but this is a way to make it more manageable.

From the syllabus: Participation: 5% Homework (5 assignments) : 25% Kaggle Competition: 30% Final Project: 40%

Long, painful weekly problem sets (20%). Two midterms (20% each) and a final (40%), fairly graded and easier than the homework problems. You'll have to work here.

30% - Five problem sets (they lie to you and say four, calling the first one "Problem Set #0"). 30% - A competition on a website called Kaggle introduced by a really hot guest lecturer dude that was actually pretty fun but for which no formal instruction that could help students do well was given. 40% - A huge project where your grade is determined sort of randomly based on a five minute midterm talk, seven minute final presentation and how much the instructors like your "data product," a term that is never really clearly defined. I say randomly because no instructions were ever given on this assignment that were not changed at a later date. Also somehow they added a participation grade in based on attendance at some point in the semester, but as of press time I have no idea where that goes.

weekly problem sets, 2 midterms, final. allowed to bring cheat sheets with equations. she needs to make the exams a bit shorter, but she's fair.