November 15, 2013

Sigman, Karl Silver_nugget
[IEOR E4404] Simulation and IEOR 4106

Please keep in mind that this review is more than 5 years old.

I have to say, CULPA completely let me down on this one, as I was completely disappointed with this instructor. Some people might say he's a nice guy, but it was obvious that he really didn't care about teaching either of these classes, and for that reason this class was a major disappointment. While his typed lecture notes were excellent, this is not an adequate substitute for lecturing.

Lectures generally started 10 minutes after the scheduled start time and ended at least 20 minutes early. I think there was one lecture the entire semester that ended at the scheduled finish time. There were several lectures that ended after as little as fifteen minutes, which seemed like a complete joke. If I'm going to go through the trouble of making the trip to lecture, I prefer to sit there for a full-length 75 minutes. I really don't feel like I learned much at all in these classes, and the level of detail was very shallow. I know we could've covered so much more if he were motivated to teach these courses, which is a shame because I could tell that there is so much interesting material in these subjects.

Homework assignments were generally rather simplistic and formulaic, with an occasional tricky one. That being said, if you didn't know how to do them, you really could just go to recitation and the TA did all the problems. The exams were in a similar fashion, with relatively few questions. The great majority of the questions were very straightforward, and there were a couple tricky ones that basically set the curve for the class. Partial credit was minimal to none, and most students disputed their grades after receiving their returned exams. Because the regrade process was completely unregulated, the success of these disputes seemed to be a function of how persistent you were and whether you had any mutual friends with the TA who graded the problem. In my own experience, there was a time that I only received points back for a problem after asking the TA twice.

The fact that the exams were open notes also seemed to give many people an unfair advantage, as some people had resources that professor probably didn't expect people to have. There were also a couple questions on the exams that were literally copied from the homework or the notes, and this seemed really stupid and pointless for an open note exam.

Despite the fact that the courses covered very little material and the exams were very straightforward, that didn't mean it was easy to get a good grade. The amount of cheating in these classes was basically insanity, and it was obvious to everyone in the class. Therefore, the averages were very high and the curve was rather harsh. For example, my average in simulation was a 90% for the semester, and I received an A-. Seeing as there were such few questions on the exams and the exams were so heavily weighted, this meant that the grading system was very unforgiving.

TL;DR These classes are simplistic in the material covered and and you probably won't learn very much. That being said, there's a good chance you won't get a great grade either.

Workload:

Simulation:
15% Homework Assignments
40% Midterm Exam
45% Final Exam

Stochastic Models:
10% Homework Assignments
45% Midterm Exam
45% Final Exam

January 29, 2013

Kou, Steven
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

Professor Kou was a very clear and easy to understand professor. Simulation was a hard class, but he had an incredibly lenient curve - if you were half a standard deviation above or below the average you got an A-. SERIOUSLY WHO DOES THAT IN SEAS???!?

He posts lecture notes and basically just goes through them word for word in class on the blackboard. This makes it easy to follow along and to study later on.

Workload:

Weekly problem sets, and roughly-weekly (sometimes twice a week) in-class homeworks (these are not announced, so you have to actually attend, or at least make it to class 15 mins before the end), 2 midterms, a final

January 18, 2012

Blanchet, Jose
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

I'm an undergraduate student so please keep that in mind as I review this class.

This is an awful class. It's incredibly stressful for someone who has had only one course in Probability & Statistics, no experience in Matlab and very limited recollection of ODE and Calculus tricks. This was the same for all undergrads in the course.

Nothing that Jose said or wrote made any sense. The lectures were way over my head but a few of the Asian graduate students followed along and had very lively debates with Jose. It's next to impossible to teach this course to undergrads and grads at the same time especially grad students who have strong backgrounds in probability.

Workload:

The homeworks were super tough. Luckily, a few of my friends had the homeworks from the previous semester and that's how we got by.

The midterm and final were impossible. I got a 48/75 on the midterm (grad and undergrad average of 64) and a 43/100 on the final...and I got a B+. But I easily could have gotten a C-. The grading is incredibly arbitrary.

December 15, 2011

Blanchet, Jose
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

I'm writing this review before my perspective of this class is colored by the final exam and my grade.

The good:

1) Blanchet is highly intelligent and he seems to be working at a different level than most people.

2) He's fairly nice and he genuinely cares about this class. He's willing to make changes if he feels that students will learn more, but has absolutely no compassion for absences or lack of motivation.

The Bad:

1) He can't teach very well. From what I've heard, he's improved significantly from the previous years. However, he moves too quickly and digs into the material before he outlines what it's used for. Sometimes, his review lectures are more useful for picking up the material.

2) The homework assignments involving programming have little to do with the examinations, which makes them tedious if you're just trying to pass the class.

3) Blanchet enjoys challenging his students. He makes practice exams challenging, but then makes the actual exam almost exactly the same. That means that this class ends up being execution more so than problem solving. Though graduate students may not be that smart, they are much more precise than I was.

Conclusion:

I would avoid Blanchet if I could. I feel that Mariana does a better job of conveying the information and adding relevance to the subject matter. Jose's class has felt like a random assortment of topics rather than a class to introduce statistical computing and simulation. This class shouldn't be a Ph. D seminar, let's cover the basics and explain why the topic is important. Let's not focus half the class on efficiency and reducing running time.

Workload:

7 Bi-weekly problem sets, easy programming in the beginning requiring basic matlab and implementation of equations, hard programming at the end requiring problem solving and actual implementation

Midterm was fairly straightforward and almost exactly the same to the practice. However, the material is presented to be much more daunting than it actually is and undergrads bombed the course.

May 18, 2010

Olvera-Cravioto, Mariana
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

If you got to be God for a day, I bet you couldn't create a being as rude, obnoxious and condescending as this professor. The TAs (John Zheng, Juan Li and Yiping Du) had the same attitude - "we are Ph.D.s, - undergrads are stupid and the cash cow MS students even stupider, so don't waste our time on doing the TA work for we are being paid because they don't deserve it."

One gem includes telling us on Thursday, the last of the reading days, at 6:00 pm, that if you want your 2nd mid-term back you have only 1 chance to pick it up - tomorrow Friday 4-6 pm.
The worst is how they took out the grade-book on course-works once the finals were done so that you never found out how much you got on the final. As far as I am concerned, this is what they did - due to graduations etc they had to submit grades by Friday night, 2 days after the exam - since the TAs grade those and Ph.D.s generally don't have exams this was doable, but why bother doing an all-nighter for these bunch of loser students so lets just put in the same score as their current average as the final exam score, assign the grade and not bother grading the finals, or do it eventually and make sure they get exactly what we put in earlier (given that grading in this course is completely arbitrary this is not very difficult to accomplish). I know that sounds impossible, but I don't put it beyond this teaching staff to do it.

The shame is that this is probably the single most useful course to have taken when you end up on the job-market. So, do your best to avoid this professor and teaching staff.

Workload:

20% Weekly homeworks - will be assigned rain or shine including right before and right after exams. Graded by graders (completely arbitrary). Officially, there are "extra credit" problems, but until today I don't know how they factored into the grade (does the extra credit count only towards that homework, or does it count towards homework in general?) . They are due at "6pm" on Fridays outside the professors door, but since slum-dog MS students will steal the stapler if they are let into the department with no one to supervise, your id card won't open the main door after 5 pm, so effectively, it's due at 5pm for you if you are an MS student.

25% 1st mid-term. Extremely long, not by accident, but by design. This will be after a deceptively short practice mid-term is posted on course-works, so don't be fooled by the practice. Basically a bunch of algorithm to simulate this or that questions which have at least 10 answers, but unless you do it the way she's thinking you won't be able to do the remaining sub-parts which build on the algorithm from subpart a) of the problem.

25% - 2nd take home midterm. If you want an A range grade in the class it's imperative that you cheat on the on this "48 hours" mid-term by working in groups although you are supposed to work individually. The way it works is that you will be given a bunch of incorrect code written by the professor (implementing things about Markov chains that you haven't yet seen in class so you can't just go correct it yourself) and incompletely specified questions and these will be modified in the ensuing 24 hours when people send her emails with questions, so you effectively have only 24 hours to do the supposedly 48 hour exam. If you work alone, you will run out of time, but if you cheat by working in a group, while you waste time trying to make incorrect code handed to you spit out the numbers, you groupmates will solve the other problems so you won't run out of time. If you don't cheat then you will end up with a high 80's score and get a B/B+, the cheaters will get 90s and A's. No action will be taken against the cheated.

30% - Final. not long and straight forward, but probably never graded, so don't rely on it to get you across the post.

In all cases, grading is either completely arbitrary or non-existent. Don't expect 1 meaningful feedback on anything.

May 13, 2010

Olvera-Cravioto, Mariana
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

Mariana isn't really as nasty as the other reviews make her out to be. Though she does say some strange things in class eg smth like ' if you thought this class will be easy, you're wrong', from her grading system and course structure, it doesn't seem like she's really out to punish students.

The exams are quite reasonable- if you work through the assignments (for which the TAs provide more than enough help for), it should be okay.

Her lecture notes are also VERY structured and clear- perhaps even the best structured amongst the courses at the IEOR department I've done, which I really appreciated. I found that I didn't actually need to use the course textbook as all the required information was in the notes.

Not sure about her lecture style as I didn't actually manage to make it to very many of the 9.10 am lectures. But from the few that I went to, and from comments by some friends, she seems like a pretty clear, well prepared lecturer.

Workload:

Weekly assignments. Programming involved can be time consuming at first, but gets easier as the semester goes along. 2nd takehome midterm was quite awful, but very generous grading done for that.

May 06, 2010

Olvera-Cravioto, Mariana
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

WORST PROFESSOR IVE EVER HAD BY FAR. Incredibly rude to students, condescending when asked questions, and basically just a cold heartless zombie. Simulation is a hard class to begin with, and hard classes come with questions - of which she will answer none because she feels you "should already know the answer".
Stay away. for the fucking life of you - STAY AWAY. This can be impossible if you are ieor, as the department must have decided that Mariana is awesome enough that she should be the sole teacher of this class-- but if for some reason there is another option - take it. you'll hate yourself, and your life, and everything you once loved if Mariana is in your class schedule. trust me.
The only plus side is she posts her lecture online, making it unnecessary to have to go to class and see the miracle that is a body working without a heart.

P.S. she just had a baby, which is what a lot of students have been attributing her meanness to, but I had her for a different class, before the baby, and she was exactly the same way.

Workload:

Time consuming, difficult weekly problem sets, unfair in class first midterm, long take home second midterm with programming. in class final.

May 13, 2009

Blanchet, Jose
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

Let me start off with the 2 main points of this review:

1) Professor Blanchet is a brilliant and accomplished engineer and is in general a nice and sometimes funny person.
2) Professor Blanchet is the worst teacher I have ever had at Columbia, in terms of both difficulty of material and lack of teaching acumen.

Basically, you will be expected to have retained everything you have learned in ODE, probstat, stochastic, and any other math courses you have taken prior to Simulation to do well in his class. Professor Blanchet's English abilities are poor at best, and this is fine in the case of some other professors (for example, David Yao of Financial Engineering E4700); however, his language skills are poor to a point where the material isn't being correctly conveyed, and most students either doze off or leave the class in confusion. He is also very easily distracted and often goes on a wild tangent and departs from the material he is supposed to be presenting.

The midterm was very tough, especially for undergraduates. About 2/3 of the undergraduate class dropped the course after the midterm and in general I'd say a little less than half the class dropped after the midterm. The final is even harder as he opts to test on material that hasn't been taugh in the class in what he considers an "educational opportunity" to learn new simulation techniques. Wonderful.

My sincere word of advice: if you can, wait until Maria Cravioto teaches the class again. If you are like me and had no choice but to take it with Jose because of scheduling...welcome to club misery.

Workload:

7 homework sets, can drop 1; last 2 took about 10-15 hours to figure out and program in Matlab. Midterm 30%, final 45%, both ridiculous for most undergraduates who aren't math Olympiad material.

January 15, 2007

Olvera-Cravioto, Mariana
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

Prof. Olvera was great. I thought Simulation was going to be a nightmare but Mariana made it much bearable. Class starts off relatively easy but gets pretty complicated after a while. You should attend most of the classes and at least a few of the recitations if you want to do well. A background in programming helps a little but more importantly you need to have a solid understanding of probability as she spends little time reviewing.
Everything before the midterm is a lot easier than everything afterwards so study hard for the midterm and get an average score. After that you should be fine for the course as long as you do all your home works and do a good job on your project. The final is a bit tricky but if you managed to attend most of the classes or review the lecture notes weekly and remain on top of the material you should be fine. Luckily she posts her notes online so you always have notes handy even if you never went to class.

Workload:

Weekly homework, some can be tedious and extremely aggravating if you have forgotten your probability. Midterm was easy, project wasn't bad at all, and the final was fair. Curve was awesome.

July 21, 2006

Sigman, Karl Silver_nugget
[IEOR E4404] Simulation

Please keep in mind that this review is more than 5 years old.

Classes were recorded in CVN, though it's definitely worth it to show up to class because questions get asked that aren't picked up on the video. He's very thorough with teaching, explaining things several ways and with concrete examples. He will drill a topic into your head with repetition. His voice tends to go monotnous though, so drowsiness is a common symptom.

Workload:

There were 3 quizzes, one of which was dropped. Homework was easy for the most part. Midterm was difficult but the final was easy, exactly like the practice final he gave.

Directory Data

Dept/Subj Directory Course Professor Year Semester Time Section