Data Analytics in Business

Dr. Xu's lectures are pretty much gold. They're structured really well with clear planning, he has amazing delivery, and he explains the concepts really well. I would say that these lectures are clearly better than Goldsman's and Sokol's, which are generally considered the gold standard in OMSA. He also does give clear direction for several avenues in these lectures as well for the ambitious student to further their learning beyond the required HW. However, I just think that the course is overall a bit too easy. I think the course would be much better if it were just a bit more difficult, e.g. if the assignments weren't so similar to his coding tutorials, if there were even some quizzes, etc. I think I would have gotten a lot more out of this class if some of the graded work required me to struggle a bit more with the concepts

(Actual workload: 15 minutes / week) The hardest part about this class is remembering that you're enrolled in it. It kind of feels like just a way softer version of ISYE-6501. If you have any background with programming and data, and I might even include taking ISYE-6501 in that, then I would recommend pairing this course with another unless you want to do a bunch of extra self-learning as well. My wife and I had a baby during the semester that I took it alone so it worked out perfect for me.

Your grade is 100% homework, and the homework code is essentially handed to you. I spent about 30 minutes on each homework and assignments are every other week. Of course if you're not familiar with R or the models in question then you'll need to spend more time, and if you don't have a newborn then you'll probably feel the obligation to do some extra self-learning. Either way, use this course however you need to cause I don't think it gets easier than this.

I was initially hesitant about taking this course because of the negative feedback on its lack of structure and usefulness, but I also could not run away from it since it's a compulsory module.

That said, I was fortunate enough to take the revamped version, and I’d say I actually enjoyed it. The good thing is that the weekly homework is essentially a direct 'copy and paste' of the lecture exercises, with a few adjustments to the data and variables. As long as you grasp the lecture content, you’ll do fine. This also becomes a bad thing, as the course does become a little too straightforward for graduate-level.

Still, as one of my first two modules in OMSA, it was a good way to ease back into studying without being overwhelmed by coding and troubleshooting on my own. Taken alongside ISYE6501, this course felt like a simpler, more structured version, though there was definitely room for the homework to be more challenging/stimulating.

TLDR: Engaging, straightforward, and an easy “A". Seems much better than the previous version of MGT6203.

As part of the first cohort to experience the updated course materials, I found this class to be well-structured and thoughtfully designed. While many of the models will be familiar to those with a background in Data Science or Statistics, this was the first time some of them truly clicked for me—more so than in other courses.

Here’s how I typically approached the weekly work:

(1) Video Lectures: The weekly videos averaged about 90 minutes (ranging from 60 to 120 minutes). I usually spent 3–4 hours watching them slowly, pausing frequently to ensure I fully understood the content and to look up any difficult concepts.

(1.1) The final video each week was a demonstration of an R exercise. I always replicated the entire exercise on my own machine, which helped reinforce my understanding of what the code was doing and why.

(2) Assignments: After watching the videos and completing the R exercise, the assignments typically took 1–2 hours. They closely followed the class demonstrations, so you could often reuse much of the code structure presented in the videos. 2 assignments are due every 14 days, you could technically do both at the same time the second week, but I highly recommend doing 1 per week.

(3) Quizzes: These included both numerical questions (based on assignment outputs) and reasoning questions. The final few questions in each quiz were usually more conceptual and required careful reading and reflection. I spent about 1 hour on each quiz.

(4) Piazza: I spent about 30 minutes per week reviewing threads. The level of Piazza activity was moderate but certainly not overwhelming.

One of the most valuable tools I used during this class was ChatGPT. It helped me understand many of the mathematical equations and concepts that would have previously taken hours to research or ask about. Used well, AI tools like this have enormous potential in education, especially for exploring examples, getting clarifications, and breaking down complex ideas. I feel a lot more empowered now to take algebra-heavy classes after this experience.

Quizzes were open-book and untimed, which removed a lot of pressure and allowed me to focus more on learning than on time constraints.

Final Thoughts: This class can be as challenging or as manageable as you make it. If you aim to deeply understand every formula and statistical derivation -like I did-, it can be demanding. But if you already have a strong background or focus more on the conceptual understanding and takeaways, the workload is lighter.

TL;DR - Five Stars. Incredible course, low workload, easy A, and good option to pair with a heavier course if you're already experienced. Take early in your program, it's HIGHLY applicable to pretty much any DS role.

I'm wrapping up the new DAB with Dr. Lizhen Xu this semester. I don't have any experience with the previous version for comparison, but based on what I've heard about it many people complained that they felt like it didn't teach useful material and that the group project was a PITA. I heard it was just something you had to get through. I'm happy to report that this new revamped course with a new instructor is nothing like that. The new version with Dr. Xu has been an absolute pleasure and I'm sad with how quickly it was over. For context, I'm a full-time data scientist with a decade of experience. While I didn't learn much from this course, I found it FAR superior than other courses and material introducing these topics that I saw when I was starting out and/or what my organization currently uses to train analysts.

Structure: This course is taught in R and in a Spring/Fall semester you have homework due every two weeks that is split into two parts. There are recorded lectures to watch that comprehensively prepare you for the assignments. You complete the homework offline and then take a quiz. There are no group projects nor any large proctored exams. The workload is pretty light if you already have a bit of experience with R and data analysis. Other than watching the lectures during my commute & gym time, I've spent 1-2 hours a week on this course. If you're a newbie, you can expect to spend ~10 hours per week according to reports from classmates. Each module is a different topic. The course starts with linear regression, then into some more nuanced types of regression/classification models, then into unsupervised learning and text mining, and then finishes with a fun baby introduction into neural networks and deep learning (you will review the code for a functioning Convolutional Neural Network that you can run if your hardware is sufficient). This class builds naturally from 6501 both in content and technology used, but you could take it as your first course in the program as Dr. Xu builds up your R knowledge from zero. While the NN/DNN content isn't deep enough to really give you a new skill, the rest of the topics are covered well and should immediately result in new tools in your toolbox. If you've been in the career-field for awhile, you will still probably pick up some tidbits you didn't know before.

Homework and Quizzes: Each homework is completed offline and then you take a non-proctored quiz that asks you two types of questions. The first type of questions are concept based and pretty much free. They ask you to interpret code or terms, fill in code snippets, or interpret the results of your analysis. The second type of questions require you to correctly identify the output of a certain step of the homework. These can be tricky due to both the large attention to detail required to not select decoy wrong answers, but also because they are entirely dependent on completing the assignment correctly at earlier steps. Errors made early in your code can carry forward and cause you to get multiple wrong answers later. In the future, I think the homework should be done in a Vocareum notebook like iCDA so that students can test their variables before moving on. Currently, there's no easy way to tell if your output is wrong other than not finding one of the four possible answers in your output. The first homework is worth 8% of your grade and the rest are worth 13% each.

Lectures: These are prerecorded and split into two types. The first is the presentation of the material via voiced over power point slides that are common in this program/format. They are excellent. Dr. Xu structures the information well both in terms of how the slides are designed and the amount of content, review, and pacing. While this in an intro course, he does a good job including deeper details like the formulas for distributions and models to keep advanced students interested. He has a slight accent, but not enough to cause anyone troubles understanding him, especially since he is one of the best oral presenters I've encountered in school or my career. Dr. Xu not only clearly communicates the content of each module, his tone and pace keep the lectures interesting. The second type of lecture is Dr. Xu coding and analyzing data similar to the assigned homework. While you can't get the exact answers for the quiz , you can clearly see step-by-step how to accomplish every task along with the exact code he used. Just like the formal lectures, these videos are well narrated as Dr. Xu explains what he is doing and why. These are good enough to maybe even take the crown from Dave Goldsman as top lecturer in the program.

Office Hours: I haven't been able to attend most of these, but I've watched the recordings of most of them. First, the man himself held office hours every week. You actually get to directly interact with the instructor! This was a welcome change from other courses I've taken. Dr. Xu is not only incredibly helpful during these sessions, but quite personable and entertaining. The TAs host a homework review/prep every week as well. Just like every other course I've taken in the program so far, the TAs do a fantastic job helping students out and providing tips and tricks.

Complaints: Like I mentioned above, I wish there was a built-in mechanism for students to check their output before the quiz and identify their errors. I didn't personally have issues, but many of the newer students did and then you have to walk the fine line between helping them with their code and academic integrity violations. My only other complaint is that this course was so light. While I recognize this in an introductory course, I wish it wasn't. I feel like 6501 already does a fine job introducing people to OMSA , and I would prefer that Dr. Xu add another 80%-120% of material so that this mandatory course is more useful to already practicing students. Everyone will have their own opinion on this obviously, and I recognize that I'm wishing that the course was something different than what's it's designed to be.

Conclusion: I can't possibly recommend this course enough both to novices and adepts. It's easy, engaging, and most importantly, useful. I suggest taking this as early in your degree as possible, and consider pairing it with another course as the workload is so incredibly low.

I was the first cohort to take the revamped 6203 and I would say that this course is an extremely good course. The workload was light but I've learned A LOT. It teaches you the basics of everything.

The course covers everything from as basic classification to regression to more advanced topics like text mining, deep learning, and neural networks. The only deliverables are homeworks which makes the approach to learning low risk.

I think this course is a fundamental course that everyone should take at the beginning of the OMSA journey. This course will definitely prepare you to tackle more challenging courses in OMSA such as regression, CDA, and simulation.

10/10 would recommend

It feels like this course was originally designed as an elective for the business school, but for some reason, it never made it through as planned. To justify the effort and resources already spent, it seems like the material was repackaged and made into a mandatory requirement for analytics students.

Unfortunately, the result is a course that doesn’t feel well-aligned with the needs of an analytics curriculum. While it touches on some business concepts, they lack depth and practical application. Meanwhile, the analytics material is overly basic and redundant compared to what’s covered in other courses.

This class could benefit from a clearer purpose. Either it should focus more heavily on business-related content to serve as a proper introduction to that domain, or it should be moved to an elective track where students with specific business interests can choose to take it.

This class feels underwhelming and lacks a clear purpose as a required course. The material is overly simplistic and doesn’t add much value for students pursuing analytics or data science. Here's a breakdown:

Cons:

Oversimplified Content: The business concepts covered are too basic to provide meaningful insights or real-world applicability. Redundant Analytics Material: The analytics topics are a watered-down version of material that is addressed more comprehensively in other courses, making this class feel unnecessary for most students. Lack of Focus: As a required course, this class doesn’t justify its place in the curriculum. It would be better suited as an elective for the business track, allowing students to take additional statistics or computer science electives that align with their goals. Pro:

Grade Ease: The class is an easy A, which is one of its few redeeming qualities. Suggestions for Improvement: To make this course worthwhile, it should either double down on business concepts to provide more depth and practical applications or be moved to an elective track for students pursuing a business focus. Without these changes, it feels like a missed opportunity for a more impactful learning experience.

This class was easy, and definitely not much of a burden even with the project, but after I completed the course I asked myself why? I really don't understand why this course is required and should at most be an elective for the business analytics students.

As for the material, the first 5 weeks covers regression, which is already covered in other courses. If anything, having a full class in regression be mandatory makes more sense than whatever this course was (I understand there's gripes with the regression course as well but I just think a mandatory regression course in general makes more sense for the program than this class). The professor also simply reads off the slides in the videos so they're really boring and unengaging. The next 3 weeks are in finance and although I found the material interesting and well taught, it was very high level and may make more sense to have an elective course on it that goes in greater detail. Then there was marketing which was just some of the lowest quality lectures you may experience. The videos consist of long tangents about topics such as the history of marketing and making a Facebook Ad campaign. Really not meaningful to an analytics degree. Fortunately, as of this semester, there's no closed book testing on this section so you can get by without watching the videos and just flipping through the notes to answer any questions on them. Lastly there was supply chain which again I thought was well taught but is 1. already taught in mgt 8803 so it's not necessary here, and 2. Is pretty high level and I think may be better off as an elective.

As for the grading in this class, it's really a joke. There are multiple choice self assessments which you can pretty much just flip through your notes to answer the questions without even truly learning. There are 3 homeworks and half of it is pretty much just like the self assessment, and then the other have is coding based which they provide similar examples for. The coding portion is probably the easiest to mess up on in this class but it wasn't difficult as far as a grad degree is concerned. For the midterm, the multiple choice was closed book besides a cheat sheet but it was mainly on regression and some finance so fortunately it didn't involve studying the marketing section, and then the coding was open book multiple choice questions. For the final however, it was completely open book and untimed which was nice for the easy A but I again didn't feel like I was really forcing myself to learn the material.

I didn't find the project to be much of a burden as long as you get a good group at the start of the course, chip away at it throughout the semester, and follow the guidelines. I've heard they don't offer it every semester anymore so this may or may not apply.

I really don't understand why this course is mandatory and what we're supposed to get out of it . Sure it's easier and a lighter course load so you can pair it with something or take it alone if you want a light semester, but is that what a master's degree is about? I've seen others mention that the course being lighter allowed them to really focus on learning R which can be true but again, there are free online courses you can take if you really want to learn that in depth. I really think it would be more worthwhile to eliminate this course and make a different course that's more valuable to this degree mandatory, because I couldn't tell you what I gained from taking this

This course reminded me of an afternoon workshop I had in middle school where we learned about 'business'. That content we covered that afternoon was probably the same depth and difficulty as this course. This course is such a colossal waste of credits. It made me regret doing OMSA instead of OMSCS. The course is a high-level overview of some disjointed business topics (including regression?!), which is not why I signed up to do an analytics degree.

Anyway, the finance and operations sections weren't too bad. It just all felt disconnected and too high-level to be worth anything. One of the 'readings' was a 3-page Vanguard brochure. Like, seriously? Is this graduate school?

By far the worst part was the project though. First of all, I was shocked by how many people actually didn't have a github account. They claim that employers and former students have told them it was the most valuable part of the course, but I've seen nothing like that on OMS Central. All I could think about while doing the product was, 'If you worked for me, I would have fired you a long time ago', but instead, most of the time I had to bite my tongue and say something constructive about the GPT generated code instead while I spent a few Saturdays rewriting the report. Seriously, people were pushing code with comments like # replace with the actual filepath, # Assuming you have computed xyz. What a horrible class. I think without a team I could have done the project in a day, with a team it probably took me a few weekends. Way longer than the combined time I spent on the other parts.

My advice for prospective students: take it over the summer - They drop the project. My request for Georgia Tech - Please make this course optional. At least for the Computational and Analytics tracks.