Simulation

I must say, this class took me by surprise. It was my 7th in the OMSA program and the first where I earned a B (all A’s until now).

Like other students, I found Dr. Goldman’s style and humor made the lessons genuinely enjoyable. A big plus.

That said, there’s room for improvement. First, I wish the Arena lessons were updated to reflect the current interface. Using recordings from years ago, when Arena looked very different, was both confusing and frustrating. Honestly, it might be worth replacing Arena entirely with a more relevant workplace tool (e.g., Python). The Arena material felt caught in between: too advanced for a simple intro/visualization, yet too basic to make us proficient. Even now, I couldn’t build more than the simplest models. I also disliked that the exams included questions requiring memorization of Arena-specific syntax, concepts or naming. In today’s world, this information is seconds away with a quick search, so testing recall over application felt outdated.

The exams were also more challenging than expected. We could bring a limited-size cheat sheet, but many questions relied on very specific formulas: if they weren’t on your sheet, you were out of luck. Again, it felt more like a test of memorization (or cramming as many formulas as possible in your cheat sheet) than understanding, which I find less relevant in a real-world context.

I certainly refreshed and strengthened my math, statistics, probability, algebra, and calculus skills (and would even say that I look forward to doing more in these fuilds!). However, as someone noted on Piazza near the end, I felt there was a lack of a holistic, “big picture” view tying it all together. The sheer volume of content (especially in the shorter summer term) may have been a factor, but I would have liked more emphasis on connecting concepts across topics. This was the first OMSA class where complex demonstrations weren’t just shown in the videos for you to follow along. Instead, you actually had to work them out yourself -pen and paper in hand- calculating antiderivatives to find probabilities under the curve.

Overall, Dr. Goldman is an exceptional teacher with great presence and delivery. But the course content could benefit from updates, a more modernized toolset, and stronger integration of concepts into a coherent whole.

Took this in spring 2024

The Professor is such a dedicated and passionate guy but this class is essentially a stats/probability 101 and I was underwhelmed by the rigor and depth, not to mention the irrelevance of some of the topic to what I was expecting. If you have math/stats background, this will be a walk in the park but keep in mind the tedious weekly homework (I got so sick of it lol) If you don’t have math/stats background and want to freshen up on some of that, this course is for you.

Final grade: A

CS undergrad and took discrete math, calc, stats 1 and stats 2, so most of the math felt pretty basic. The early modules especially (like 1 through 4) just go over concepts that probably should’ve been prerequisites. A placement test at the start could help filter that out and let the course go into more advanced content sooner.

Later modules (7–10) were much more interesting. Unit 7 was a standout—I had no idea that taking the ratio of two standard normal variables, like those generated via the Box-Muller transform, produces a standard Cauchy distribution, which helps explain the undefined variance of said distribution. There were some real gems like that I’ll be telling people at parties;). You could tell the professor was really into it, especially in the second half. He kept saying things like “this is my favorite topic,” and you could genuinely feel his excitement.

The exams were kind of underwhelming. Most of it felt like high school-level stats problems, with a few random Arena simulation questions tossed in that didn’t really connect well to the rest of the material. Like how is it important to know what shape this particular button is, or whether function X can be found in spreadsheet A or not.. And for all the focus on calculator rules (like half of the questions TAs answer are about them), I honestly think every exam could’ve been done without one. Maybe in the future try no calculators or just the honorlock/desmos one. If you also provide basic formulas and ban cheat sheets, I bet TA workload would drop by 25%.

Finally, instead of adding arena trivia to exams, turn one of them into a proctored simulation project? Something open-book, but no LLMs allowed, where you have to create and analyze a couple simple simulations similar to those in the lectures. That would feel way more relevant and give people a chance to apply what they’re learning.

Overall, the professor himself is a solid 5 out of 5. Might actually be the best one in omscs I've had so far.

First, I think this course is a great course to take along with CSE 6742 (Military Gaming) since that course goes over internal model validity, while this course is more nitty gritty. You might never need to write a pseudo-random generator from scratch, but it's important to understand how the foundation of your tools work.

Aside from that, the professor mentioned their intention to refresh the course and to remove the arena specific grading and I agree - the arena modules are nothing but correctly memorizing the very specific questions ("Oh you memorized the Advanced panel (which isn't even in the Comp Lab version of Arena)? Sorry, this question was for the Basic panel"). The group project I think is a nice balance of learning and not a tremendous importance to your grade, but I think the large curve in the class speaks a bit to how rigorous the math sections are - I think some of the math questions are little bit unfairly transformed from the problems given in the questions and practice exams necessitating the curve. Lower the bar a little bit to more align with the problems as given in the lectures and I think the curve wouldn't be as necessary. Outside of that, the cheatsheets present a great opportunity to study - but you'll also spend a tremendous amount of time trying to cramp every little bit in. I also thought the final was a little bit too broad - since its cumulative, there's a lot of problems in the last two modules I would have expected to be problems (a lot of "ah... I bet this Antithetical Random Numbers portion will be a test question") only for them to not be problems and to have to remember where on my cheatsheet I put parts about Newton's Method. Dave's enthusiasm and the TA's helpfulness on Piazza really made me enjoy the class and I believe I learned a lot, but it is by no means a perfect class

If you are looking for a course about using simulation in an operations environment, this course is terrible. If you are looking for a graduate level statistics course that is depressingly called Simulation, then this is the right course.

I will never use any of the materials in a professional setting ever. For example, there is a one week module on Pseudo Random Number Generators. Is anyone going to write one from scratch? No, you will just use the rand function. If you want random numbers from a different distribution, you will use that function.

There are a few weeks spent on Arena (again a piece of software I will never use in a professional setting) but it more a how to use the software sort of module.

80% of the grade comes from two midterms and a final exam. They are timed multiple choice without any materials except for a one page cheat sheet. Since you cannot use any outside tools or materials this is a big game of memorization. There are alos series of questions about in what drop down menu in Arena certain functions are contained . . . .

There is almost no time spent trying to connect the stats work to actual simulations and how to evaluate simulations. You are asked questions around calculating expected value from functions that are transformed with equations, but never is it explained why you would ever do this in a professional setting.

If you can't tell, I did not like this course.

Adding a data point on grade curves for this course. My final grade in Canvas was 83.83% and my final grade was an A:

Exam 1: 85 / 100 Mean: 78.22 Median: 79 High: 100 Upper Quartile: 88 Low: 36 Lower Quartile: 70

Exam 2: 67.5 / 100 Mean: 75.5 Median: 77.5 High: 100 Upper Quartile: 85 Low: 25 Lower Quartile: 67.5

Final Exam: 79 / 100 Mean: 77.16 Median: 79 High: 100 Upper Quartile: 88 Low: 37 Lower Quartile: 67

Project score: 90 / 90 Mean: 83.45 Median: 89 High: 95 Upper Quartile: 90 Lower Quartile: 85

I found this course to be the most challenging in the program. Granted, I was on the B-track and am no good at statistics. Dave Goldsman is a gem, though, and made this course the most engaging that it could be.

Extremely rewarding stats class as someone with little formal experience in statistics or math during undergrad! I did this course simultaneously with AICS.

I loved Prof Goldsman's energy, enthusiasm and corny humour in his lectures, and how he eased us into the math as someone with math anxiety (he was always saying "don't panic! I just copied this formula from the last slide" etc.) Definitely kept me going when work and life outside of OMSCS was challenging.

Weekly MCQ homework assignments are an easy win, but the 3 timed, open book (cheat sheets only) exams take up most of the weightage and time spent. The exams are challenging and preparation was time consuming (spent about 2 - 3 working days preparing per exam and got about 80% each time), but because Prof Goldsman made the content self-contained and threw in a bunch of easy questions every exam, I felt that I was assessed fairly relative to the effort I spent preparing.

The only slight complaint is the focus on Arena for 3 weeks of the course - wasn't sure about the utility of the content, when there are easier (and more transferrable) methods to do simulation through Python or general programming languages.

Even though I'm no math whiz by any standards, overall, I left the course feeling much more well educated in stats and more confident in my own math skills. Whatever grade I do get, it would be definitely well-earned. Thanks Prof Goldsman (and teaching team!)

I took this class this summer and if I were to go back and take it again, I'd say make sure you do the practice exams with you cheat sheet and under time constraints. If you struggle with a practice exam question, make a note of it, and mark it down on your cheat sheet. Every so often I see myself revising the cheat sheet all again and again due to the feedback I got when doing the practice finals. If you think that 2 practice exams isn't enough, I'd go and look up textbooks that cover the same math and stats content as what is being taught in Sim. (Make sure it covers RVs, Prob Dist, etc) and do those questions. Honestly Sim is just a math course with applications using ARENA. Also,

I'd recommend getting a good scientific calculator and leverage their stat's functions there. I personally bought the FX 991 ES and it helped so much because it had integrals and you can switch modes where you can calculate binomial, normal, and Poisson (just look at the calculator guide and see how to do it). The project is lenient as long as you follow the rubric. We picked a topic where it felt like a research paper than anything (we did literature review, methods, results, discussion . Any additional requirements we list it in the annexes). But overall I'd say your main focus for this course is the practice exams and whatever question you can get ahold of using other probability/stats textbooks.

Good think about Dave is that he knows how to teach without reading the script so its more engaging then, lets say ISYE 6501 where Dr Sokol reads off the slides a lot (no hard feelings to Sokol, but I gotta give aura points to Dave for not reading off the slides). I never really had to look for external resources beyond Module 2 and everything is pretty much self contained as he said.

Its a great class overall, it prepares you for more advanced classes like CDA. Definitely a well structured and well run course!

I initially didn't want to take this course mainly because I used SimPy in 6501 instead of Arena because of how many complaints other student had with Arena. However, the reviews for PM and DO weren't as great at this course and a friend convinced me that it's really not a lot of Arena. She also repeated over and over how funny Dr Dave Goldsman is. I'm sure that in my long career as a student, I haven't met a funnier professor... he would have excelled as a profession comedian if he would have pursued that path! I was surprised at the immense amount of material covered in this class. And, as usual at Georgia Tech, they found an infinite number of ways to generate problems on the material. If you have time, I recommend going over the bootcamp course... It should help you with the first exam, at least. Watching the posted exam review questions were really helpful as well... Although I caught on to those late. I felt comfortable doing the HW problems by myself but, a word of caution, they can bring a set of totally different questions on the exams. Make good cheat sheets and practice as much as you can. It was one of the toughest courses I've taken during the summer but fun nevertheless.