I know people have many reasons to get a Masters. If your goal is to take some higher-level classes, you can do better than GT. If you are self-motivated enough to do an online degree, you can do it for free. Aside from free offerings from sites like Coursera, you can find whole courses up from many institutions - with syllabi, lecture slides, homework assignments, etc.

If you're planning to do it for the salary, in my experience the degree ended up being worth a $95K to $115K bump in starting salary. Compare this with the 2 years of industry salary that I would have received, and the 2 years of experience (and raises that come with that). I know I wasn't paid better than the folks who had been with the company for 2 years more than me.

If you're thinking about it for the sake of your resume, I do hiring screening / interviews now (for Data Science and Software Engineering positions) - and I really don't care if the applicant has an MS or not (or what classes they may have taken). Most folks I know that do hiring think similarly. My main signal from your resume is projects you've been on and how you contributed.

From my point of view, this program is a losing proposition for any potential student.

Yes, you can study the same material on your own, but you won't earn a degree from it. Now that I've got the degree, I'm in much better shape to pursue further learning on my own.

Note, however, that I didn't do this to improve my resume, go fishing for a new job, or try to get a raise. With tuition reimbursement from my company I only spent $3500 over 2 1/2 years to earn a full-fledged master's degree.

Based on the above, I can't agree that it's a losing proposition.

Could you explain a bit more what do you mean with this statement? Is it that you feel better prepared to study advanced topics (like advanced ML/Data Science) or was the degree a requirement for something else you wanted to pursue?

I'm curious about what other "doors" having this degree opens, other than the bump in salary mentioned by others.

Congrats for completing the program btw.

I honestly don't expect it to open any additional doors for me. I'm a software developer with 30 years experience and have been working for the same company for nearly 18 years. I wasn't looking for any changes, I just wanted to be better at what I did.

Thanks for the congratulations. It was quite difficult at times, took a lot of effort, but was totally worth it, IMHO.

I think one benefit of a curated course is that it includes materials you didn't even know exists. We can easily improve on our known unknowns - just pick up a book or google it - but unknowns unknowns are... well difficult to learn. I think going through a graduate program helps you get a better grasp of what you don't know AND what you didn't know you didn't know.

I have extensive knowledge of VMs, so I helped many students get their environments set up. I could often diagnose show-stopping problems for the less-experienced students very quickly, since at my experience level I really have "seen it all". And if it wasn't something I could diagnose that way, I'd set up a Google Hangouts call and watch exactly what was happening on their screen and get them through it.

Many other students did the same thing. In Computability, Complexity, and Algorithms, there were some students who were apparently math robots from the future, solving the problem sets effortlessly, and posting them to Piazza so that the rest of us could use their work for study purposes.

I'd figure for a "light" class that had a fair amount of coding or was in a subject area that I had considerable experience would be 10-15 hours a week.

A semester with 2 classes of moderate difficulty would be 20-30 hours a week, depending on homework pacing, amount of videos and readings to study, etc.

The hardest class I took was my last class in December called Computability, Complexity, and Algorithms (CCA) and at the end I was doing 35+ hours a week trying to get ahead. It was hugely difficult due to my very weak math background, but I somehow got the hang of it and passed with a decent "B" and graduated.

That said, with a couple of notable exceptions, the graduate classes are there for PhD students as first and second year background material so they have some starting points for their research. This naturally leads to a format where the semester can effectively be described as a long reading list of papers and lectures to spur discussion on the content of the paper. I was planning on pursuing a PhD when I started into my MS, so this format worked quite well for me at the time. In the years subsequent to that, the grounding from those classes has given me starting points for deep dives into problems I encountered at work[0].

It's interesting that you brought up machine learning. Charles Isbell's Intro ML class was a significant exception to the pattern I described above. In addition to high quality, pre-prepared lectures peppered with entertaining anecdotes, the had high quality projects that worked with pratcial tooling. It was also probably the highlight of my graduate career[1].

[0]: In particular, the material covered in my graduate systems classes has been invaluable for not reinventing the wheel for the thousandth time. The material from the couple compilers classes I took on a whim has been a huge boon when talking about software correctness. I work on the hypervisor underneath GCE. Correctness is near and dear to my heart, but performance is right there with it :)

[1]: For undergrad that dubious honor has to go to Olin Shivers, not only because of his eclectic teaching style, but also because his class completely altered the way I think about problems in computer science. In particular, my mindset shifted to one of models of computation and decomposition of problems into subproblems for which the simplest model could apply. I have an example I'd like to write up, but it's a bit long for a footnote.

GT's MS in Analytics degree is actually designed specifically for people who are going to go out and work in the analytics field -- it's not a pre-PhD degree, and our courses are targeted primarily at people who want to learn and apply analytics. We have an industry advisory board that helps us target course and program content, and we're constantly working to make sure our coursework is focused to the right cohort. We even have a required applied analytics practicum (both for on-campus and online students) where our students work on analytics projects for a wide range of companies and organizations.

Perhaps other degrees are different, but the MS Analytics is a very practice-focused degree.

In general, while theory has great value, it's more as a stepping stone to higher study than as an end unto itself. Few computing pros submit proofs among their deliverables. And devising the theta bound on a function or resolving the terms of a CSP simply don't deliver much value when working outside PhD-level R&D labs and writing peer-reviewed papers.

I believe there's a great deal of value in applied non-PhD track academic programs like GT's online discount offerings, especially in serving professionals and employers. I also believe it's high time that universities clued in to the unmet need that most of us post-academics face toward helping us continuously re-educate ourselves as we progress through our careers. Few of us pros can return to campuses, even part-time. Distance learning meets a crying need. And when done right and priced-right (as I believe GT does), I have nothing but kudos to offer in return. I say, more power to GT's authors, curators, and administrators who made this possible. And to all who make this greatly empowering service possible: thanks, and keep up the good work.

If you want to play around with theoretical computer science, get your PhD. College educations are too expensive to not be imminently practical.

It's a waste of time to teach industry tools at a university. It's much more valuable to be taught fundamentals. Know your fundamentals well and any new tech will be much easier to learn. It's long-term thinking - put in the investment to make sure you can change skillsets in the future.

All the things you mentioned tend to be ephemeral and change a lot within a few years. Look at the git monoculture that's sprung up in the last 5 years for example - 10 years ago it might have been reasonable to teach SVN.

You have to do programming assignments anyway. Why wouldn't you require students to learn and use the latest source code control tools while they're doing their development?

Teach students to write tests, use source code control, utilize continuous integration, etc.

Although the specific tools, languages, and approaches will evolve in the coming years - none of the above are going away soon.

Time spent making undergraduates use git is time that could be spent teaching them long-term, fundamental skills.

there's no better way to learn something difficult than to learn it when you're young

Hmm, define 'young'. I'm in my late 20s and I find it easier to learn new things more than ever - including things I failed to learn in my teens and early 20s. Maybe I'm just a late bloomer, and it took me a while to "learn how to learn". But maybe I'm still young in the eyes of someone who has been using vi for 30 years (:

Testing ect is usually covered in all intro classes (assert libraries) or industry type testing like JUnit by a software engineering elective typically taught in Java

Just because one of gits key abstractions is based on a kind of graph, I don't think it follows that knowing graph theory means you know git. I mean LISP is based on a graph structure as well but plenty of people find that confusing.

In my experience with the MSCS program (nearly ten years ago at this point) the core required classes were mostly well structured and would serve people well continuing onto a PhD or growing their skill set for industry. The core constituted a relatively small chunk of the overall credits required, though, and the elective courses tended to be more along the lines of what I described.

I'm glad to hear that the Analytics program has a more dedicated focus on practical matters. It might be interesting to produce a series of similar (but narrower) curricula that amount to curated collections of CS classes making up degrees in Machine Learning, Systems Programming, etc.

I personally really enjoyed my dartboard-oriented approach to class registration. I learned more than I've never needed to know about approximation algorithms, cryptographic theory, and compilers. Even if much of what I learned there hasn't proven itself directly useful yet, I really enjoyed learning it for learning's sake, and I think I'd have had a hard time picking up some of the gems I pulled out of that since. I also still have a hobby of proving problems NP-complete on demand as a bit of a parlor trick (within the limited scope of problems for which you can apply the small handful of patterns I've burned into my brain over the years :).

I did, however, find that my undergraduate university had a great program for people with nearly complete degrees who had been away for a few years.

I'll be finishing undergrad this May and am now looking at grad schools. Feel free to contact me if you want to chat about this because it's been surprisingly hard to find info or advice in our situation.

Who can apply to the OMS CS degree program? Admission into the OMS CS program will require a Bachelor of Science degree in computer science from an accredited institution, or a related Bachelor of Science degree with a possible need to take and pass remedial courses. Georgia Tech will handle the degree admissions process. For more information please visit the Georgia Tech program page.

Mostly you use python numpy and scipy to analyze a large time series data set (stock market) to predict pricing while having a low correlation to the overall market movement.

I had some success and won their 6 month contest, but I still feel like a bit of a hack. I'd like to move into the financial quantitative analysis industry.

Would you say this GT program would be a good stepping stone?

What is the best way to get in touch with you and get the syllabus material for the courses?

I'm at rememberlenny at gmail.

It's amazing how you don't think of someone for almost 30 years, but you read their name in a comment on HN and memories come flooding in. What do those neurons do while they're waiting to be used again?

I also did BS and MS at GT, and while I generally share your experiences there were 3 or 4 truly disappointing classes during my MS. They didn't ruin my overall experience, but I can see how someone could happen to have more experiences like those and fewer positive ones and come aware with a very different perception of course quality.

My overall opinion of GT is mixed, but rigor or the courses is not one of my top critiques.

In the year I did it, the class was structured as follows:

At the beginning of the semester, you'd pick two datasets.

Every two weeks, you'd apply two or so algorithms that were being covered at the time (maybe k-means and SVD, or a NN and SVM) to your chosen data sets. There would be a set of variations that you were supposed to apply to each algorithm. Typically you'd normalize or clean the data in some way. Perhaps you'd filter outliers, etc...

The result would be a set of experiments to run (2 datasets) x (2 algorithms) x (2^3 variations per algorithm). You would compile the results into a (10 page max) paper, with analysis about how the dimensions differed.

It was up to the student to figure out how to actually implement this pipeline (I used sqlite + numpy/scipy/scikitlearn, many used Matlab).

On paper, this sounds like a great class - what a wonderful way to learn about how different approaches relate to each other, and how crucial the process of preparing data is to the effectiveness of the algorithm. In practice, however, this did not happen for most students I knew.

These students spent most of their time finding implementations of the algorithms and hacking at them to actually run all the experiments. They then rushed through gluing the results together through some semblance of analysis. Alumni of the class I knew said the same thing about their experience.

This analysis was read by TA's. There were I think 3 of them for about 100 students. We wouldn't get the papers back for weeks (long past we moved on to new material). When we got our papers back there was very little feedback of the content - mostly it was noted that we submitted the work on time, and had successfully performed all the experiments required.

I agree that Isbell is a joy to listen to - he is charismatic, entertaining, and I too enjoyed his anecdotes. However, I felt like you would only get something out of his lectures if you already knew what you were talking about.

When I think about the quality of the class, I think about how responsive the class is to the individual needs and progress of the student.

If you say that it's up to the student what they get out of the class, and your bar for a good class is that the content is arranged in a nice manner, then here you go https://pe.gatech.edu/sites/pe.gatech.edu/files/agendas/CS-4... ... any self-directed student can grab Mitchell, and do the weekly assignments I describe above - all for free and in the comfort of their own home.

This is especially true of term project courses, where the final portion of the project to which you devote the most time and creativity is also the part for which you're likely to receive the least feedback.

I disagree (having taken the course as an undergraduate and it being my first major exposure to machine learning). Certainly if all you do is attend the lectures, you're going to miss some background knowledge, but that is true of most (if not all) university courses. You're supposed to devote 2-3 hours of outside work for each hour of lecture. Meaning 6-9 hours of studying per week outside of those lectures.

Some of this is doing the projects, although some of it is personal investigation.

There are failings of his course (one of the biggest at this point is that it doesn't do any work with the state of the art now), but I think that the fact that his course caters toward people who are self-driven is not a failing.

The best way to look at what the goal of the course is is by looking at his exams. If they weren't different than you took them, they were intentionally too difficult for the allotted time, leading to low averages and incomplete work by the majority of students.

However, the course allows motivated students to make connections between concepts, with the help of the professor and the coursework. Having someone "leading you" down the right path is very helpful, much moreso than a textbook alone.

I really do think that there is one exam question that sums up Isbell's course perfectly: its the one where you are asked to compare and contrast 4-5 aspects of 4 randomized optimization algorithms (RHC, GA, SA, and MIMIC) and explain situations where you'd use each and why.

The course's goal is to lead to a strong intuition for the algorithms covered (sadly at the partial expense of a theoretical understanding), not everyone puts in the work to develop that understanding, but that's not a failure of the course, necessarily.

You can find the syllabus for Isbell's class and follow along. You can do the readings and programming investigations. If you like lectures, you can find many full courses on YouTube (I found caltech's lectures https://www.youtube.com/watch?v=eHsErlPJWUU to be the best at presenting SVM's out there, although this was probably my third attempt at understanding them so maybe the other resources rubbed off.. they also skim over the quadratic programming detail but I get that this may be beyond the detail that many people desire in an intro class).

If you have to teach the material to yourself, how is your experience improved by being in the class?

To be fair, most of Isbell's course (lectures) is also available on Udacity.

>If you have to teach the material to yourself, how is your experience improved by being in the class?

There are a couple advantages. One of the most obvious is the lower latency of responses when you have confusion or misunderstanding. In a lecture, you can ask a question and get an answer almost immediately. This is most useful (imo) with algorithms and mathematical concepts, because you can ask, and lecturers are often quick to provide insight, into the interrelationships between algorithms (both in Machine learning and in a more theoretical sense like computability). There are topics that come up a lot, and being able to have instant feedback on those connections allows you to spend less time misunderstanding than not.

That alone is a fairly weak justification, I think the stronger one is feedback in general. Watching lectures only gets you so far. With implementation of algorithms, often your feedback is testable correctness (although my experience in DS&A suggests that most people are capable of constructing incredibly incorrect models for things that perform well on some input, and even on decent autograders), but with things like machine learning algs and intuition about those algorithms, you can't get that. So the feedback that yes, your understanding is correct (even if that feedback is slow) is invaluable. In that regard I think online courses and MOOCs can be good, but MOOCs that don't provide feedback aren't as valuable. I've attended a lot of lectures, and I've ignored a lot of lectures. Listening to someone say something does not mean one has learned it.

I'd also note that, if I recall, the way that Isbell approaches teaching the material, vs. the way the textbook does are very different. Textbooks are (often) references. They provide information on what something is and how it works theoretically, but very often lecturers are able to provide the kinds of things that aren't (and shouldn't?) be in textbooks.

If I'm reading a textbook, its very likely that I want to know how to implement an algorithm, so I care that the algorithm for simulated annealing says that you jump with probability e^(D/T) > Rand[0,1]. Whereas in a lecture, I'm likely much more interested in the idea that simulated annealing is conceptually very similar to throwing a ping-pong ball into a large complex, convex plastic surface and seeing where it lands.

I don't agree that feedback during lecture is valuable or low-latency as you say - not with 100 students attending. It might work to ask a clarifying question here and there, but again - you're only in a position to take advantage of that if you're already comfortable with the material and are generally keeping up.

Books are different than lectures, sure, but I don't think there's much difference between attending a lecture with 100 students, or watching one online. Indeed many people claim the online way is better, since you can rewind and skip around, pause and lookup references, etc...

I'd agree that Tech has too few TAs for too many students, generally, for its graduate courses, but I don't know that other schools do a better job. A brief survey of the folks around my desk elicited howls of laughter at the notion of useful or accessible TAs in grad school.

> I agree that Isbell is a joy to listen to - he is charismatic, entertaining, and I too enjoyed his anecdotes. However, I felt like you would only get something out of his lectures if you already knew what you were talking about.

I think this assertion is, at best, too strong. A better assertion might be that his lectures depended on coming in with sufficient background.

As I said, I came into the course with no experience with machine learning at all. On the other hand, I did have a fairly strong theoretical computer science, stats, and linear algebra background. I will admit that may have made me blind to things he was simply assuming with respect to educational background that were not actually safe to assume. That said, I still refer back to his primer on information theory (http://www.cc.gatech.edu/~isbell/tutorials/InfoTheory.fm.pdf) when discussing work relying on it, so he certainly made some effort to fill in gaps as he discovered they were common.

> When I think about the quality of the class, I think about how responsive the class is to the individual needs and progress of the student.

For a graduate level course I feel a class clears this bar when it accurately and thoroughly documents the prerequisites. Now, I'm not saying Charles's class necessarily does this. As I said, I came in with a pretty strong background in what turned out to be more than sufficient, but with that background I personally felt his lectures were quite tractable, even assuming complete ignorance of ML itself.

Ironically, this sounds quite a lot like much of industry.

The question remains is if an online degree has the same credibility. Looking back at my time at GT, I cannot see how operating solo, without the constant feedback from your peers and faculty, is as good. There is more than just what is in the study material. The other question is if the entrance requirements are still as stringent.

From my perspective, unfortunately, Georgia Tech has really diluted the value of their masters in cs degree. They have become a sort of immigration visa-mill with very many India undergrad -> Georgia Tech Masters of very questionable skill level.

Just my experience.

I can't speak if the CS degree has been diluted but I will say there is an enormous amount of extremely undeserved selective bias against Indian people for technology jobs. When US citizens even see Indian names there are less likely to hire (known as name bias).

Again I can't speak for the programs but my Indian friends that went to tech were at the top of their class both in MS and BS. Highly qualified. Extremely humble, ambitious and hungry.

Just my experience (and I run a recruiting software company so I see it at scale).

I compare this to the apathetic divas that I have met from Stanford and MIT (I live in Mass so I have met many MIT grads). I would hire a GT grad over them any day.

I worked in India for one of these BPO companies and know what I'm talking about. If the good engineers from India want to reclaim their status, they need to push back against the flood of H1Bs from these companies.

But instead, the majority of HN (or at least the guys who do more hiring, less coding) keeps pushing for more and more visas when we should be urging Congress to reform the system to allow the talent to come in (with Green Card), while disallowing US companies from using it to lower wages for all US engineers.

I'm sure the hiring manager who posted earlier will say it doesn't affect the decision but the subconscious bias is a real thing.

You're doubly fucked if you have an Indian name and you were born here. White hiring managers automatically assume you're incompetent, and Indian H1B hiring managers are threatened because they fear for their job.

When I was younger, I found it odd that many of family members of mine would Americanize their name. Now, I've experienced the reality of it.

I also think a graduate degree is necessary for certain types of work. For just bog standard programming jobs where you can read a web page to learn the language/framework/library, sure you don't need it. But for other types of roles (quant roles, more research oriented, ML, anything tech cutting edge that requires theoretical understanding), an M.S. or Ph.D. is going to be a gatekeeper whether you want it to or not.

I've been out of undergrad and working in industry for close to a decade as a Software Engineer and/or Embedded Systems Engineer. I feel I'm doing pretty well in my career, but I've been looking at the online masters in CS as a way of showing that I'm dedicated to continuing learning, and to maybe open up some new doors for myself down the road.

From your perspective as a hiring manager, would this be worth the time and effort? It's not like I wouldn't be interested and learning new stuff anyways, but if I'm going to go invest the time and money to do the degree vs. learning on my own, it would be nice to know if it was worth anything.

(No sweat if you don't feel like answering, or want to take this private. Thanks!)

That said, a masters degree is not necessary, it is an overkill. I rather see open-source projects.

I see only an upside in earning a MSCS, however you do it. But while learning more principle and technique is always good, it's not strictly necessary and it's definitely not sufficient to outcompete other candidates.

Experience in relevant side-projects is a good thing. It shows initiative and passion and that you're more than just a serious student. Open source dev experience demonstrates your desire to create -- not just to design, but to actually make -- as well as your ability to work with others, especially distant others. Most pro tech work now requires not just up-to-date tools and techniques but also good communication skills, increasingly with folks who work away from you. Demonstration of such skills is unusual and desirable, especially in those just out of school.

In hindsight, looking back at my MSCS, I think the strongest point of a Masters is not the elevated agree, but the opportunity to focus on a specific field. If you have an interest in AI, go to GT and work underneath a professor with lots of experience. Do not get a Masters just for the sake of getting a Masters.

GT has tons of research dollars. I was paid the entire way there, even as a non-PhD student. At first a teaching assistant, and then a research assistant. You cannot get that kind of experience from an online school.

There are companies where side projects or off-hours work on open source projects is complicated. Employment contracts that say the employer owns all of your work, on the clock or off, aren't uncommon. It might be easier for some people to get a masters degree, especially when employers are willing to pay for it.

If staying in embedded will just focus on work experience.

From my experiences with GT, I am apprehensive about the quality of content - I assume it's coming from the same departments and professors that I had experience with.

From other comments here it seems that the approach this program took was quite different from the one employed on campus, so that apprehension may be unfounded - that is, the online offerings may be of higher quality than what students on campus receive. Still, I felt like I needed to post something to warn people that the branding of GT does not in-and-of-itself mean that the content will be of high quality. Students considering the program should try and find some way to evaluate this - are there sample classes or lectures posted online? Perhaps ask someone you trust in the industry to take a look and give you their feedback.

Still, I think students who are self-motivated should consider what they would be able to accomplish if they took some time to organize a study program for themselves. There are many free high-quality resources out there that could be used for effective self-directed study.

Students who are less confident about doing it on their own should be asking themselves what sort of support they expect to be getting from the program. Certainly there are many advantages in having things curated for you, as well as having access to discussion boards with other students going through the same material. Aside from that, many students (unfortunately, I think), need the external schedule and commitment - and for them, merely having an exam deadline, or the $10K investment looming in the background may be the thing needed to get through the material. Those students, too, should be realistic about the investment they are making and what they hope to get out of it.

>Andrew Ng's Coursera class on Machine Learning was the pedagogical highlight of my time at GT, and I did it on my own initiative (and it's free).

This is something I wanted to highlight from your post. I don't think this is surprising, nor do I think it is reasonable to state that a course (or a degree program) is poor quality because it didn't meet the standards of Ng's ML course. That is an exceedingly high bar.

Its something I noticed, because I am a recent grad, so while I was in my mid-level courses, and had recently taken Tech's intro CS course, I was able to watch (Harvard's) CS50 and other courses. But on the other hand, I've seen some very bad online courses. The successful and large online courses are successful and large specifically because they are head and shoulders better than the rest. And there are a lot of decent online courses, so to measure against what are some of the absolute best online courses is to measure against courses that have more resources, more planning, and more feedback than most.

(as an aside, they also have more incentive to be good, but that's a bit tangential to the point that they also have more opportunity to be good).

You type in "Machine Learning" on coursera and you get over 1000 results (not all of which are relevant, but assuming even 10% are), its little wonder that one or two are going to be better than the even the best courses that you'll take during a bachelors or masters, because Coursera offers more Machine Learning courses than most people will take in their Bachelors or Masters.

Combine that with these courses coming prefiltered (you've heard of the Stanford Course, but what about "Applied Text Mining in Python" from UMichigan, which for all I know might be great, but it doesn't come with the hundreds of recommendations that the Ng course does, so I don't know that it will be great) and you have a really great recipe for a bias against the in person courses.

What track were you on? Do you think that had anything to do with it? What did your peers think about the program?

A $95 to $115k bump sounds pretty darn good. Did you already have a CS undergrad degree? Where from? Sorry for all the questions, feel free to share as little or as much as you're willing.

I'd say my peers generally shared my opinion (classes not being very good). Many of them were trying to get into PhD programs so they were focusing on finding research opportunities and didn't care about the quality of classes much. Some struggled but blamed themselves for this rather than the class. (I'd say this was very common amongst the undergrads too).

Of the PhD students I knew, most were discouraged from taking classes altogether, since it took away time from research. This was true even in the first year of their program. The general attitude from that side was that classes were a waste of time.

Many higher-level classes were run as mini-research projects. You'd get some content, then the rest of the class would be forming teams, proposing project ideas, implementing, writing up results and having 'mini conferences'. I think faculty liked this since it was a good way to try out research ideas, recruit potential PhD students, and give their current students extra time to work on their research.

Nothing wrong with that format, of course, but the actual coverage of content was typically superficial. If you weren't already familiar with the area, you had to figure it out on your own as you went along. Also, this is not a class format that translates to online very well.

I do not. I'm self-taught and considering the online MS for the purpose of signaling that my skills are legitimate (and filling in some theoretical gaps). As a hiring manager, would this change the value of an MS in your eyes? Or still unimportant compared to projects?

With a non-tech BS degree, all too few HR departments (esp in bigger companies) will invite you interview for a software job. Without the CS degree, I was a pariah with very limited prospects. Frankly I doubt that POV has changed appreciably, even after 27 years. Business-men/women are a conservative lot. They take as little risk as possible. If you lack credentials, they hire you, and you fail... they look bad and have a hard time explaining why they hired you. But if you had a relevant tech degree, their ass has far better cover.

Of course, if you already have a BS in CS, I can't speak to the value of adding a MS. Even when I earned mine, the incremental added value beyond the BS wasn't overwhelming. But some employers care more for advanced degrees than others. Uncle Sam and most large companies automatically kick you into a higher salary bracket if you have one.

It also doesn't hurt if the school granting your MS is renowned. Aside from silicon valley (apparently), I suspect 95% of employers will be very positively impressed by a degree from GT. I know several employers responded favorably over the years to the mere fact I had a degree from Johns Hopkins. Like it or not, your alma mater often matters.

One problem is that MS degrees don't really cover the general curriculum. They're often, even when rigorous, used to allow students to focus on a topic or project that isn't as lengthy as a PhD. For instance, someone with a CS might be interested in numerical computing, and work on ways to solve various differential equations.

The downside here is that this means a math or physics major might get an MS in CS, and do some programming in numerical computing, but not know much about algorithms or data structures.

I'm presenting this in the context of a genuine, rigorous MS degree, because it doesn't need to be a watered-down experience to still show the pitfalls.

Some MS degrees do require certain core courses before you can apply - so they'll take a math major, but they'll require that this student complete certain undergraduate courses - some before applying, some while enrolled. This can add time to the MS degree but avoids that scenario I described above.

Of course, once you've actually taken those courses (say, a math major passes courses in data structures, algorithms, compilers, and operating systems), then the MS may not be critical for finding a job anymore. But the degree can help.

Unfortunately, I've noticed a trend toward discounting MS degrees or even holding them as a negative indicator. This is probably because people get an interview because they have an MS, but then are tested during a technical interview on general CS that they may not have taken.

Not sure of the solution. I think the best approach is to take promising students from other fields, but then make sure they've taken the additional core coursework. This would add some time to the degree, but if all MS students did this, I think the degree would be more respected.

As it stands, the BS in CS is respected, because it (if the school is accredited) must contain all those core courses that tech companies love to quiz people on.

Whether those topics are actually relevant to the job is an entirely different topic!

To signal your skills, I'd think about the industry you want to work in, and try and work on a project that is similar to work you'd like to do.

For example, if you're wanting to get into Data Science, find a data set, pick a question and answer it. Build visualisations, implement ML algorithms, etc. Put your code up online and write a blog post (or several) about the process.

A year spent doing that would be worth more in my eyes than a MS.

I'd recommend it if you have the time. As much as HN likes to push "just show your github contribution", degrees do matter to companies.

Also, hang in there. The degree may not have helped your current job as much as you'd like, but it may help more at a future company. Sadly we often need to change jobs to get a real salary bump.

The margins on adding a MS CS (or analytics) is less for a CS grad, but at 10k a degree and the flexibility of taking online they are huge for non-CS majors looking to pivot. 10k to bump your salary up 20k (while working and getting experience + raises) plus the knowledge add pays for itself.

Im biased on account of being in the program (supplementing my Electrical Engineering degree due to a change in career and life paths), and I know a lot more about graph theory, formal algorithms, and high efficiency computing than I did a year ago. These are things that help improve both your portfolio of rigorous projects (Im currently working to port over all of my high performance computing assignments to Rust) and assortment of tools for the ever annoying CS interview.

My doctorate was at Texas, and it certainly catered primarily to what brings in the $$: e.g. research.

I didn't have a Computer Science undergrad degree, so perhaps I have a different outlook on this than you did. I'm currently working as a software engineer, so I'm also not forgoing earning a living by taking the time off from school.

In other words, while I can understand why you feel disappointed by your educational experience, there are other lenses through which this program makes sense. I feel good about having gone through it so far and I'm looking forward to finishing.

However.... I think majority of people that take OMSCS (I am in my 2nd semester) and this new OMSA program do so part time while still maintaining their fulltime jobs and families.

Also, why the throwaway?

I'm getting exactly what I'm looking for out of it (a somewhat structured environment to learn in), so for me I would say this isn't a losing proposition.

Your situation is anecdotal at best. You are critiquing an in person experience to an online experience.

GT goes out of its way to say that the online experience and physical experience offer the same degree, is it not fair to compare them?

E.g. someone who did a PhD with one advisor vs another

Anecdotally I've felt all of the professors are eminently interested in interacting with MS students online.

My own experience was doing about half of an M.S. about a decade ago before dropping out to go right into work, due to money constraints. I don't regret taking those classes at all though - I learned a lot because I put a lot into it, and it's knowledge I've used throughout my career. I find it hard to believe doing an M.S. at gatech would have no value at all - I guess it depends what you plan on doing and how you will use it.

Although I agree with your sentiment (formal education has a lot of flaws), a full-fledged degree potentially solves other problems than just improving your resume to get a better/higher paying job for engineers.

Family pressure, lack of confidence that comes with not having a formal CS undergraduate degree, or motivation and structure that comes from paying for a formal program could all contribute to someone choosing this path. A $10k, online masters program from GT seems like a good option for some people.

There are no programs that have stood out to me during hiring to predict the quality of the candidate.

If you're a student looking at programs, I would prioritize the amount and quality of individual attention you stand to receive. A self-directed student can do well anywhere (including on their own). If you're not so stubborn/resilient, having a good community and mentorship to help you overcome difficult times is key.

I could have done it for 1/3-1/4 of the cost and in the same amount of time but with far less time spent commuting.

Wait... isn't this kind of a good reason to suffer through the course?

From my first job, I realized life in a cube wasn't for me. I really wanted to be in front of a classroom. I realize there are problems with academia. I know you have the same squabbles and competition you have in the corporate world, and seeking certain grants to keep yourself afloat can cut into the research you actually want to do.

Still, I really wanted to teach. I've seen so many professors who only work one or two jobs, or go straight from BS -> MS -> PhD with very little industry experience. I wanted to be a different type of professor with plenty of real work experience to drawn on and teach from.

Grades don't matter. I've found that's very true for industry. Having a GPA on your CV doesn't really mean anything and most people leave it off. However, it has a huge impact on getting into degree programs.

I only had a 2.5 undergrad and even though I got a 3.2 in grad school, it wasn't enough for most programs I looked at. I attempted and failed to get into 8 schools back in 2009 (ironically, one that I later worked for and could get free classes at. PhD programs however, are full-time).

Today I have three publications that I'm 2nd author on, and in 2015 I attempted to get into school once again. I contacted several professors. Most simply don't write you back, but even when I got in touch with several schools, many simply didn't have any professors who were willing to take students in my field (environmental sensor research).

It's really competitive to get back into school and there is a massive disconnect right now between industry and academia.

You get out of any education what you put into it. You can leave with just a basic understanding of computer science and only know two languages leaving an outstanding program. You can also go to a crap program and push yourself to learn more on your own; using what professors teach as a jumping board for a lot more.

The TL;DR I'm getting at is that masters programs do have a purpose: getting you into a PhD program. If your work pays for it, it might be worthwhile for the additional title, but if not, you're not going to learn anything you couldn't apply yourself to on your own.

I took a master's in aerospace engineering at a top tier school and I wonder if this is the norm for grad engineering courses, and for many lower-level undergrad courses as well (think massive freshman calculus lectures). To calibrate what you consider "poor quality," how did you find the quality of your undergrad engineering courses?

Sure the top students in the program are going to do well, by definition, but there are plenty of more "middling" people like myself that can only be brought up to the next level with proper discussion/interaction with classmates. From my experience even PHPbb would be a more effective tool than Piazza.

---

Suggestions (if Piazza folks are reading):

1. Allow one to delete follow-ups.

2. Allow students to create private "study group"-like threads that aren't in the main feed.

3. Make it easier to upload pictures and other content.

4. Make things live. Normally this wouldn't be necessary, but anyone in the program knows many students post the same thing at the same time as a response to an event (like an email). By doing this you prevent redundant threads from being created.

5. Use some sort of up/down voting system that way the community can self-regulate.

There are plenty more things I'd improve, but for the sake of brevity those are some I just came up with on the spot.

Disclaimer: software engineer at edX, have worked on discussions features in the past. Happy to answer questions.

Overall, IMHO Piazza clearly has far more features.

That said, I still prefer edX forums simply because I hardly ever need any of the Piazza features - and the Piazza GUI IMHO is much worse than edX especially for the target group. After edX implemented the sort-option "by activity" and "show only unread" my major complaints were solved.

--

By the way, since I hope you read this reply, the worst by far on edX is that the darn page just won't stand still!

Please stop the "convenience scrolling (by Javascript)". Also, the height of the page changes when I select a comment that does/doesn't fit in the viewport (long threads vs. no-responses-yet comments, for example, and that too triggers automatic scrolling.

Next, the size of the text box changes! I use Chrome to make it much larger to see my entire long and carefully crafted and formatted comment, then I switch to another tab to do more research for my reply, but when I click back into the textbox it shrinks back to its small default size. Could you please leave my manually-made-large textbox as it is?

No changes of scroll positions, no changes of sizes, no changes of anything, please. Not just in the forum, the course pages have the same annoying (anti-)"feature". I find it very inconvenient when the GUI changes under me, it's the GUI equivalent of walking on ice.

I would also enjoy a feature to "mark all as read", so that I can use "show only unread" at any point in time. Right now that feature forces me to have clicked on each and every comment before it is useful.

Life update would be "nice to have" too, especially for a TA (if it's too expensive for server-load, how about offering it just for the TA and STAFF user groups).

What is also missing is even the slightest hint (in the comment editor, or anywhere) that you support LaTeX style formulas! As well as some help for beginners how to use it.

I know Christmas was last month, but that's my wishlist :-)

What I don't miss are votes. That always gets misused, and it is of very limited benefit. I see the value of votes in forums on something like reddit, I don't see benefit in edX forums with a focus on Q&A. Especially not down-votes. Even the "pin to top" doesn't really work: At least 70% of users don't ever notice pinned posts. If there have to be votes, do it like Disqus: You can see who voted. Or instead of votes do what Github did in the issues comments.

By the way, why do you still have votes at all? They are only displayed if anyone specifically checks them in a given comment. They used to be shown in the comment overview on the left. Right now the "vote" feature can be removed and nobody would notice.

Some random ideas:

Feature: Allow Staff and TAs to post without their status. Just like on reddit. Not everything I had to say was "official", and one reason I turned down several invitations to be a CTA on edX was because I really hated to have a green banner around every single one of my posts. Too much pressure, and just because I'm TA doesn't make all my answers "precious".

Feature: Allow admins (TAs, Staff) to merge threads. For example, the introductions at the beginning where 500 people post their own "glad to be here" thread instead of using the pinned(!) "Post your introduction here" thread, or when there is an issue and 50 people each report it (even when the last 20 new threads are about that same subject, most people post immediately before looking at the forum).

Feature: Detect if you responded to this person in the past. It's nice and gives a feeling of familiarity to find out you talked to someone in the past.

Feature: Geolocation. Show where the other people (in the current thread) are located.

Feature: Combine Wiki and forum: From within the forum let people select comments and/or threads for the Wiki. Ask the for some additional information, like choosing a Wiki page/section. This can address the problem that good comments are quickly buried by all the new ones.

Bug report: It seems right now the "Section" or "Chapter" or whatever you call it is not displayed in comments.

[1] http://umichdesign.com/

If the same issues creep into the class, we'll reconsider Slack. What probably won't change is the popularity of online collaboration tools for all workers. Fostering a community of similar learners to ask questions, share work, and help each other is one of our goals for the semester.

The more I'm out in the working world and interacting with people from a diversity of educational backgrounds, the more I agree with that approach.

(I'm specifically thinking about the "No, we as a profession tried that in the 70s, and here's why we decided it was a bad idea" moments)

Maybe this condition does not hold true for the OMSCS, but I think almost all your problems can be solved by a good moderation team.

My experience so far has been excellent. I just started my second semester, and I can say that the curriculum covers exactly what I wanted to learn with the exception of one class. The program is extremely practical, it's only one year and is focused on getting the students jobs. The professors are great, and I highly recommend it to anyone wanting to get into the field.

Most of the core classes (like machine learning) are more math based. In machine learning, it's done from a mathematical/theoretical approach. Another example is regression analysis, where you learn the math before doing the practical work in R.

I have a weak math/stats background but a strong software background, which also aligns with my interests so I'm doing the computational track. As I mentioned before, most of the core curriculum is math heavy so I'm learning that aspect, but I'm also getting a ton of practical experience in machine learning and big data work.

Looks like the business track isn't offered in the online program (yet), which is weird because it looks like all the requirements are offered online.

- CS7450: Information Visualization. Basically learned how to create useful data visualizations. Loved this class

- CSE 6242: Data and Visual Analytics. This class is interesting because you learn a very wide breadth of tools. We learned visualization (D3), big data processing (hadoop), and analytic tools (random forests for example). This was my favorite class so far

- ISYE 6414: Regression Analysis. The theoretical math parts of this class were really hard for me, but learning how to do practical regression analysis in R was super valuable to me.

- ISYE 6644: Simulation. Loved the professor for this class (Dr. Goldsman) but the material didn't seem very useful to me. Knowing the general approach of simulation for modeling is useful, but we went pretty deep into the math that I don't think will be useful in the long run

- ISYE 8803: Intro to Analytics. This was the most practical class I took and was taught by the head of the program. You basically learn all of the analytical models for the first 2/3s of the class, then in the last third you look at case studies and how to apply them.

Hopefully there's enough detail in there for you, let me know if you have any other questions

Thanks for all of these comments, great to have some inside knowledge on this program.

Most (I'm guessing 75%, possibly higher) have some work experience. Most have either strong math background or strong software background but there are exceptions. Lots of former engineers of all kinds, chemical, mechanical, software, etc.

"Will the degree I receive from the OMS CS program be the same as the on-campus MS in Computer Science or will my degree say “Online”? Your diploma will read "Master of Science in Computer Science," exactly the same as those of on-campus graduates. There will be no "online" designation for the degrees of OMS CS graduates."

https://www.omscs.gatech.edu/prospective-students/faq

---

An employer could simply notice that your location and your school are in different places. However, IMO it's more impressive to have been working a full time job and doing a degree than the opposite.

AFAIK, this is Georgia Tech's second online masters degree course, which demonstrates that the university now stands behind their online degree program. One of my coworkers is currently enrolled in their other online masters program, and he speaks highly of it.

If this is anything like their first online masters program, there is no distinction between the online and on-campus diploma. The degree is a full-fledged Georgia Tech degree. This is what makes it so attractive--employers and other educational institutes won't know you didn't physically attend the campus.

I'd be happy to get more of my coworker's opinion on the program if you like. I'm interested in it myself...

I'm a current student and I switched jobs about halfway through the program. I'm living in Austin so it came up in the interviews that this was an online program. I explained to people that I was doing the online degree in my spare time but taking it slower and that all the courses were the same as the on campus counterparts. Generally speaking the interviewers seemed to be impressed by this and said that it showed a strong work ethic.

I can't give hard facts about the value of the program but given that it only costs between $7k and $10k to complete I think it will pay for itself very quickly and perhaps already has.

1. https://twimlai.com/twiml-talk-4-charles-isbell-interactive-...

And everyone who's been exposed to both says it's the same courses and rigor as the on-campus version.

(Source: 2 classes from finishing)

I am curious: who in the world would do both?

A more serious question, since you're going through it now: is there a thesis component to the online MS, or is it just coursework? If there's a thesis (either required or as an option for more research-oriented students), how does the thesis mentoring at a distance work?

Nobody, they're TAs now. ;)

Got any info for the lazy on the existence of a thesis track?

In my experience, the thesis is a huge part of the degree when it's present so I find it at least somewhat curious that some (many?) programs that are generally considered quite good don't have one.

Another thing one professor explained to me is that the standards for a Master's thesis are actively changing. You can either do a glorified undergrad senior project (but with you doing all of the work that a small team might do in undergrad) or intense publication-worthy research. The latter was mostly people going forward with a PhD to follow. I definitely went in to the degree with the impression that all Master's students were doing a thesis in the second category, but I think this is a case of the times are changing.

My BS CS was also from GT (several years before), so much of the core course list available online were slightly more rigorous retreads of distributed algorithms, simulation architectures, etc. already covered late in undergrad. The most valuable courses in my MS were special topics courses that, at least at the time, weren't available online, many of which were in other departments as part of the computation + application domain interdisciplinary approach of the program.

That said, the core courses that were also offered online were generally quite good, and a few were exceptionally great. Rich Vuduc's HPC applications course comes to mind.

So based on what I've seen and read, I could recommend it.

- Although not universally true, you tend not to need labs stocked with equipment that aren't practical for an individual to purchase. (Though, yes, CS degrees can be oriented around hardware.)

- Automated grading of problem sets/tests seem to work better with software than just about any other topic of MOOCs I've seen. (Though that's probably not an issue for this sort of paid program as you can have paid TAs and professors directly involved.)

- Remote "teams" can clearly collaborate effectively on projects etc. as demonstrated by the fact that they do in many companies and on many open source projects. That said, a big part of a masters degree is going to be individual work anyway.

ADDED: As someone else mentioned, in-person mentoring would still be a concern of mine but that's probably a tractable issue.

http://www.londoninternational.ac.uk/

Three of my kids and I are are doing CS degrees from here right now, two of us while working full-time.

https://www.coursera.org/university-programs/masters-in-comp...

(1) Understanding statistics. Hopefully this program will take care of this requirement, but it's not hard to find these people anyway. There is an infinite supply of science PhDs fleeing the academic job market.

(2) Behavioural/personality. People who will do well at the actual job. Example: can you tell when a PM is asking you to answer the wrong question, and how do you handle it?

You can easily find (1) with screening questions, (2) is the hard part.

But, I guess if you think you have (2) as a future analyst, this program could be a good way of getting (1).

http://www.wgu.edu/online_it_degrees/data_management_analyti...

Is there any business in creating a better "classroom" experience that what ex. Piazza is doing?

It seems like an are which could be improved immensely design/ux wise but it also seems like it could be an area where that's not really going to make you successful because the distribution is already owned by someone else.

I joined the program because I come from a non-CS background - undergrad in math, work in an unrelated field: consulting. I'm trying to pivot into a ML Engineer career. If you want to learn ML, you're better off going through the Ng Coursera course and from there pursuing some personal projects. The primary value of the program is the ability to get past recruiting coordinators simply due to the fact I'm enrolled in CS program.

The two undergrad CS courses I took at Berkeley were more rigorous, and were superior from a skills development perspective. But at the price, the OMSCS program is definitely worth it for someone coming from a different background.

As someone with a BS in Materials Science & Engineering (at best a tangentially related field via sparse EE coursework) who does some level of programming at a tech job now, I'm curious what my prospects for admission would be. I'm confident I could handle the coursework, provided I could get my foot in the door.

As a related question, they mention taking courses to fill holes - are they receptive to Coursera offerings?

And yes, online learning and/or self-taught learning is definitely fine as hole-fillers; we just want you to have the necessary background to succeed, and however you get that background is up to you.

It's not just a course it's a degree. When you graduate you have a Masters degree that is identical on paper to an on campus degree.

Because even if you're unemployed, your most valuable resource is your time.

*Assuming there's enough in your savings account to cover expenses

However, if you have the motivation then it's certainly do-able. My work was fairly intense too so as long as you are comfortable with giving up your weekends and have supportive friends/partners then it works just fine in my opinion. You get used to studying as your "fun time" - eg when I commute I spend it reading course notes, I play next to no PS4 games these days and I watch very little TV. I did think about taking a work break to totally focus on the course but in the end I've not needed to and find I prefer the brain ping-pong.

As someone who was in a relationship for 5½ years and is now single if you have a partner do talk it through with them. Be as supportive as you can of them on your journey. I'm pretty sure one of the reasons I lost her was the intensity of the last year or two. That was definitely exasperated by the workload and my general need to dedicate my weekends to exams/coursework/studying.

It's like running a side project that's actually used and not just for learning/experimentation.

Also: Even at two classes per semester it takes two years. One class per semester, a bit over three years. It does take some determination to see it through to the end.

(Either way, if you are going to teach programming at all version control is a big part of programming so it should be taught.)

I am not misrepresenting you, I m not really even arguing with you or disagreeing with you all that much.

I agree, online degrees are not providing job skills. But neither are many (not all) traditional degrees. Just one example is people fresh out of school who have never used or even heard of version control. Then there's the "can't do fizzbuzz" example.

This isn't even exclusive to CS degrees.

The reason I choose my degree program is it advertised itself as career focused education.

Either we should make education more career focused or accept that universities aren't primarily for job training and find alternatives.

Are there any similar tracks that do not have this requirement?

https://pe.gatech.edu/online-masters-degrees/analytics/faqs#...

--EDIT-- On the admissions criteria page, it states the following:

"5. Optional - Applicants may choose to submit standardized test scores, most commonly GRE or GMAT (but if appropriate, LSAT, MCAT, etc. scores may also be considered)."

So that may help get you in.

https://pe.gatech.edu/online-masters-degrees/analytics/appli...

I think both schools offer a track that emphasizes AI/ML, though I'm not sure more than a couple of their available courses are in those topics.

Provided they do not relax their standards, I don't see there being an issue. An MS appeals to professionals who are unlikely to relocate for a program. I'm in a suburb of NYC; if I wanted an MS, I'd have to consider commuting to Princeton or Yale, at an hour and a half to two hours each way, to the much closer NYU or Columbia but with similar travel times once the intersection of mass transit schedules and class times are considered, or to the state school 15 minutes away with a much less prestigious program.

Meanwhile, GT has a prestigious program at a cost comparable to or better than the state school, with no commuting issues. It's the program I'd go into, and I live close to a surfeit of prestigious CS programs! There are plenty of qualified applicants across the country that don't have the option of "drive to your favorite of the 3 nearby Ivies".

If they keep the admissions process and curriculum/evaluation equivalent between online and in-person, then going online greatly expands their applicant pool but doesn't necessarily dilute it. If the program decides to relax their standards to get even more money, then they'll have problems, but that strikes me as being penny-wise and pound-foolish.

Also has a lot to do with selectivity. At the entry level, companies can rely on selective institutions to do some of the vetting for them.

One is a Udacity Certification, the other is an Accredited Georgia Tech Masters Degree.

"Udacity is not an accredited institution and we do not directly provide college credit. We have, however, partnered with Georgia Tech to offer an accredited, fully online Master’s Degree in Computer Science. While the courses are hosted on Udacity, the degree is conferred by Georgia Tech. Learn more about our Georgia Tech partnership"

https://udacity.zendesk.com/hc/en-us/articles/207991913-Can-...

Here's the link to it:

https://www.appliedmathonline.uw.edu

Georgia Tech is especially strong in operations research.

So, here data science is a new bottle of wine blended from some now quite well known old bottles of wine. And it is not nearly the first such blending since there have also been programs such as mathematical sciences and applied mathematics. Other blendings have included mathematical finance, financial engineering, and bio-statistics.

Apparently the high current interest is because now the associated computing is much cheaper, more powerful, and easier to use. And there has been a lot of hype from some sources.

However, I question if US mainline business is much interested: IMHO and my experience says that nearly any specialized technical material faces a serious obstacle since in the organization chart the highest ranking technical person (if not the CEO then necessarily a subordinate) has to report to a supervisor who knows from much less to nearly nothing about what that technical subordinate person is doing.

MD doctors, CPA accountants, licensed engineers, and licensed lawyers have some crucial, serious professional status, processes, support, etc. that is missing with applied mathematicians, statisticians, data scientists, etc.

For software developers, roughly, the solution is for the organization to have a CIO, all the developers are in the CIO's organization so report only to experienced developers, and only the CIO reports to, interfaces with, non-experts in computing.

Computing is now so darned important that the rest of the C-suite has to swallow their pride and accept the CIO at the table.

Net, I fear that data scientists will have too little professional or organizational protection from rain falling down the organization chart from the C-suite.

Or, for the supervisor, most projects will be lose-lose: If the project fails, then the supervisor has a black mark from wasting money on a failed project. So, with a failed project, the supervisor loses.

If the project is successful, then the supervisor and, maybe, everyone in the C-suite, maybe even including the CEO, can be afraid of the project leader now regarded as a 900 pound tiger and, thus, a loss for the C-suite.

Here the organization chart from the project leader up to the CEO is engaging in classic goal subordination, that is, pursuing what is best for themselves personally while sacrificing what is good for the company.

And for startups, what fraction of venture partners would be able to evaluate a proposal that makes heavy use of some of the more advanced applied math in that Georgia Tech program? Net, the venture partners don't know the technical material, either.

Or, as I suggested, nearly all wine in the blend is now quite old, and it didn't achieve much traction in mainline business.

My short summary view is that for such technical material, especially material more advanced than in the Georgia Tech program, and for a startup, the founder CEO needs to be both (A) the main expert in the technical material and (B) essentially a solo founder who can write the software, bring it to market, and get the coveted traction significantly high and growing rapidly -- at which time the founder may not be willing to accept equity funding and report to a BoD that does not understand the work, that is, be back in the situation of a technical subordinate reporting to a supervisor who does not understand the technical work and, with the low expenses of a one person company, just grow organically from revenue.

Or, IMHO, the most promising career future of an applied mathematician, etc., in business is to be a solo founder of a startup.

Edit: There is placement data for the on-campus program: http://www.analytics.gatech.edu/placement. 95% within 3 months of graduation, cohort size 21 students, with the majority (40%) taking an "analyst" title, average salary $100,000 (61% going to Atlanta, so this could be a little depressed compared to west coast tech salaries).