IEOR E3106

1 midterm (3 problems), 1 final (~8 problems), weekly HW (half of answers from recitation, half given in class by Kou the day they're due)

Problem sets each week with about 6 questions - mostly from Sheldon Ross Probability Models Textbook

Stochastic models is probably one of the most interesting topics you could learn but Ward manages to be such an incompetent teacher that this is essentially a self-study course. The class covers, Discrete and Continuous Time Markov Chains, Poisson and Exponential processes (not distributions - there's a big difference) in detail, Renewal Theory and Brownian motion. All very interesting topics that can be used to model a variety of scenarios and in general improve your understanding of probability. The problem is not the material, but Ward. He doesn't follow a well-defined thought process, jumping around between topics, hastily scribbling something on the board and erasing it before we can comprehend what he's trying to do. Attending lecture was a total waste of time. He often admitted that he knew that we had no idea was going on but we should all read the notes posted online (beware - the notes for Renewal theory stink so learn it some other way). The TAs go over HW solutions in recitation before they are due. They're a bit harder than what you see in the exam most of the time. First midterm was ridiculously easy with a 91 median. Second midterm was ridiculously hard with a 56 median. Final was on the harder side since there's so much to study, but the median was in the high 70s I believe. I feel like his exams have gotten harder over time given previous years tests. He has a…

The Good: -Gives out copies of lecture notes before class -Lets you hand in HW AFTER he goes over half of the problems -Tests based entirely on problems from HW, class, and sample tests -Very approachable for questions after class -Extra long office hours before tests -Curves to an A- -Lectures flow well; he is well prepared -Takes some lectures to teach application of Stoch (ie. stock pricing, options pricing) The Bad: -Accent sometimes hard to follow; sometimes stutters -The subject material is hard--you won't understand it even after the final -No lecture notes if you don't show up to class -In class quizzes can be difficult if you didn't pay attention Overall: probably the best teacher you could have for Stoch in the IEOR department. I've had Sigman for ProbStat before, and I consider Kou an easier A and an equally well organized teacher. The paper lecture notes he hands out are a huge help, though he doesn't lecture as well as Sigman.

For years I had heard about the notorious Antonius Dieker and how this class (an already pretty difficult class in its own right) was going to be absolutely miserable with him. Coming into the class with this in mind, I was quite surprised to find that my experience was the complete opposite. I think this is arguably the best class I have taken at Columbia and Dieker was arguably one of the best professors I have had here. There is no doubt the Dieker is pretty brilliant and that comes through in the way he teaches, as does his passion for the material. He never made light of questions asked in class and always insisted on people asking questions if there was confusion. The midterm exam was surely not easy but I would by lying if I said it was an unfair exam. 10/10 would recommend this class

Weekly Problem Sets solutions given at recitation 2 Midterms 1 Final

I don't care what anyone says, I love Ward Whitt. The material is somewhat challenging and averages tend to be low around the 60-70s, but if you put effort into learning on your own, you will enjoy it. If you don't you probably shouldn't be in IEOR. He posts all the lectures and previous exams online. It will definitely pay off working through several to prepare for exams. Know the definitions of distributions and different stochastic processes and everything will fall into place. "Well I watch more TV now than I did in college... (thinks for a second)... but that's probably because we didn't have TVs back when I was in college." -Ward Whitt

11 weekly problem sets, gone over by TAs, graded by completion 2 midterms Cumulative final

This is the worst professor I have had at Columbia. Many other students I talked to shared a similar experience. Those I talked to considered themselves better off by NOT going to class, to save themselves the confusion and grief. His lectures are completely unorganized. He skips all the important steps and never explains anything. He rarely if ever covers any kind of general case. He's all over the place in terms of explaining anything. He has a bad habit of writing the crucial step or equation on the board, but then immediately erases it because "that's not important." It's incredibly frustrating. For how impossible the material is, however, the class is run okay. He has his TAs go over the problem sets before they were due (though the TAs weren't much better at explaining things). Even then, the problem sets are graded on completion. He puts plenty of practice exams online, and goes over them in class and in TA sessions. The material you see on exams is usually easier than what you see on homework. The tests are open book (closed notes). Unfortunately the book is horrible. The grading is pretty lenient. He splits grading into parts: 1 part for all homework assignments, 1 part for each midterm, and 2 parts for the final. Each part is scaled against the class average: for example, if you have an 80 on an exam, and the average is a 60, then you'll have a 1.333 for that part. Y…

1 problem set per week, but everyone has the solutions, and the TA's give out the answers during recitation. 2 Midterms, each very similar to past year midterms 1 Final, very similar to past year finals