# courseProbability

 Dec 2010

On the first day, she started by saying that probability is hard and that it took her until after she received her PhD to truly understand it (which is really an indication of her lack of ability and of her condescending attitude. more on that later). She also told us that we shouldn't expect to understand probability in just a semester. That's a great way to start a class right? Lecture: She teaches right out of the book (which is available free on the internet as a pdf. So don't buy the textbook if you are fine with printing out the book). This means that if you are fine with reading from the textbook, then you will gain almost no extra understanding from attending class. Also, don't expect her to have a great handle on basic math concepts either. For example, she doesn't understand basic calculus. If you see the way she does double integrals, you will understand what I mean. Homework: She assigns a homework on Wednesday and it is due the following Wednesday. This means you will get plenty of practice. On the first day she promised us that she will write up all the problems in the homework and that we won't be able to find solutions. Don't worry. That is a lie. Most of her problems are copied verbatim from similar courses at other universities. So, with creative googling, you will probably find solutions to most problems. Exams: The exams are fair and open book/open notes. They usually take most students the whole class time. If you did the homework and know your way around the textbook and your notes, then you should be fine. Just a tip: print out the wikipedia pages for the different distributions like binomial, poisson, etc. the pages have some useful proofs and lists of useful expressions like PMF, MGF, etc. TL;DR: the class is manageable, and the professor is horrible. Find solutions to the HW on the web. Print out wiki and bring notes and book to exam. Lecture is straight out of textbook.

 Dec 2008

Very Hard class. make no mistake about it, this is a very hard class. It says its for people who have never seen probability, but its really not. If you haven't taken a class in probability before, don't take this class expecting to coast. Mariana is a good professor who follows lectures very closely to the textbook. The course starts off easy, but gets very hard in the middle section. By the end, this was a big time commitment with homework, which would have been impossible without the recitation where the TA gave you the answers(go to recitation!) Overall, tough tough class with a good professor.

 Apr 2007

Professor Cravioto is straightforward. Her lectures follow very closely to the text about 70% of the time. The thing is, her lectures are very concise and she tells you all the information you need to know. The homework assignments arent too taxing but there is usually a problem or two that will take some thought, especially if like me you had no exposure to probability/statistics before. I recommend attending recitations because TAs clarify somethings you may not get from her lectures. The first midterm was insanely easy, the second (almost in response) was way harder with a fair number of people failing it, and the final was fair and not too hard at all.

 Feb 2007

Possibly the worst class I have ever taken at Columbia. She was new this year and seemed not to realize that the problems sets should not be exceedingly difficult for an introductory probability course. The exams were straightforward.

 Jan 2006

 Dec 2005

Professor Coffman is not a good teacher. It is hard to get around that fact. Students catch on early in the course and by the time the first midterm rolls around attendance is hovering around 20%. I don't think I have taken a class at Columbia where the average student put in less effort. The exams were all straight forward applications of the material in the book and the class. He himself characterized the final as a "feel good exam". Nevertheless the averages scores were around a 50%, making it very easy to get a good grade. The class itself covers an interesting set of topics though it does tend to move slowly, and you are held to a very low standard.

 Dec 2005

Professor Coffman is a very nice man with an incredible knowledge of probability theory and its applications. Sadly, he has great difficulty communicating that knowledge. His lectures are generally unsystematic and difficult to follow, and he makes frequent conceptual and arithmetic errors which make understanding the material based on his lectures alone something of a game of probability in and of itself. For the first two-thirds of the semester there were weekly problem sets, but after the class average on the second midterm came in at about 30% I think Professor Coffman gave up grading them. Now he distributes optional problems from a similar course at MIT. If you have an option, do not take probability with Professor Coffman. If you do not have an option and only care about your grade, you can probably get at least a B simply by showing up to class (and exams) and doing the problem sets. If you really want to learn probability and can only take Coffman's course, get another probability text to supplement Bertsekas and Tsitsiklis and prepare yourself--from the beginning of the semester--to spend several hours a week reading from B&T _and_ your own book and doing problems that are not required. Probability is an incredibly interesting and useful branch of mathematics that is notoriously difficult to teach. It is unfortunate but ultimately unsurprising that Professor Coffman has come up with no revolutionary methods for making the study of the field less painful.

 Jan 2004

RUN RUN RUN and DON'T TURN BACK! This class was one of the worst academic experience of my life. Professor Coffman is a great guy, but one would be better off not going to his lectures. Not only are these lectures boring, but they are filled with a plethora of mistakes that bring one's confusion a step further. The terrible book does not help, either. One is tricked into believing that Coffman is an easy grader on the first day, but DON'T BE FOOLED! His grading rue brick is vague, and in the end, he just arbitrarily affixes a grade to your transcript. In addition, most exam problems are just embarrassingly hard compared to the homework. AVOID THIS CLASS AT ALL COSTS!

 Dec 2003

Prof Coffman is a renowned engineer who has spent 20 years in Bell Lab. He's a very nice old man whose office hour is 24/7. He would give you his home no. and invite you to his house to discuss problems. Excellent approachability. He would try his best to pull up the grade of the whole class. But he devoted all his free time to office hours because the average score of the test was as low as 40%. Not too many people understand what he is teaching in class and the textbook he uses is terrible --nothing more than a whole bunch of equations written by some MIT professors. nothing to worry about the term grade, as i said, he will give you make-up exams, extra credit hws ..etc. he's one of the most devoted teacher i've ever seen, but his lectures are not as great as his personality.