Current OMSCS grad student; three down, seven to go. Loving the program so far.
The content is great, and most of it is available on Open Courseware, YT, etc, but here's what else you get by officially going through the program:
- the amazing community of TAs
- the assignments
- the feedback on reports & projects (either automated, or through TAs)
- the collaboration with other students on Ed, Discord, Slack, etc
- the forcing function of deadlines, having to study for exams, etc
- free access to academic libraries, IEEE, ACM, O'Reilly, etc
- access to software and services, educational packages from GitHub, Wolfram, Google Colab Pro, student discount in a bunch of places, etc
Another underrated aspect is GT's ability to preserve rigor of the program overall, despite the scale and number of students in some courses (the most popular ones have 1,000-1,500 students per semester).
If you're on the fence on applying, I strongly recommend you do. The program is affordable enough that there's no harm in trying for a few semesters to see if matches what you're looking for.
Graduated 5 years ago. One of the best decisions in life. Coursework is challenging and more than 2 classes would definitely feel like a very full time degree (1 per semester is the best pace imo for work life study balance). Although you would need special permission to take a third class (from what I remember).
My resume is also looked at differently after mentioning Georgia Tech. It really helped gain a lot of confidence. Fundamentally changed things as my undergraduate in India was not a good experience for me.
In reference to the open courseware, is there a way to either just download all of the videos in bulk, or view them as part of a single video? It looks like they're broken down into ~2 minute long video clips through the Ed platform, which is very annoying.
Annoying indeed. I created a script using ffmpeg to merge all the 2-min clips into a longer video per chapter[1], so I could watch the lectures on my commute.
You may need to tweak for different courses, but I've used for ML4T, GIOS, and ML, and it has been incredibly helpful.
I wish! I travel quite a bit for work, so it breaks my legs every time it happens. Plus family, kids activities, etc. ML was brutal this semester, but hoping the curve will help a bit.
But it's ok, slow and steady is the way to go. Besides, I'm doing this for the fun of it; I don't need the diploma for career or anything.
OMSCS requires ten courses to graduate. I completed one course (with an A grade) before realizing that, even at a pace of one course per semester, it was not a high enough priority for me to devote the time required to do each course well.
That course was great, though, and I definitely learned some things I'm glad to have learned!
IMO the instructional materials are a small part of the value. The things that stood out to me were:
- the assignments
- the autograding of programming assignments
- giving and receiving peer feedback about written assignments
- learning some LaTeX for those assignments
- having an artificial reason (course grade) to persist in improving my algorithm and code [on the problems taught in that course, I wouldn't have been self-motivated enough if they were just things I came across during a random weekend]
The ability of OMSCS to scale paper writing, review, and grading with real human TAs is nothing short of astounding. While it's a ton of work (I'm just completing class #5) it's a great resource for both learning the material - and how to communicate it effectively.
I have taken three of those classes as part of the Online Master of Cybersecurity program. They were all excellent. I can say that the assignments were an important part of the learning experience, for instance the practical experience of attacking weak RSA keys.
I would not let the lack of assignments, tests, and quizzes stop you from trying these if you are interested. At a minimum, they would give you a feeling for what the program/s are like, and possibly encourage you to enroll into the online degree program, which is an exceptional value.
I’m an OMSCS grad - the dedication to making higher education in CS more accessible is something that really sticks out to me from those in charge (shoutout to Dr. Joyner who heads the program). Although not every course is on the Open Courseware (nor course work), there’s still a lot of good material, and if you like it enough, the program is a nice little side quest in ones journey through computer science.
Hi is there a way to view the lectures in a more traditional way? For example, as one long video? I'm seeing lessons broken up into 2-5 minute long videos.
I once considered applying, but I gave up because collecting letters of recommendation was a major hurdle. My academic advisor from university has already retired…
Get recommendations from supervisors you've had. Academic references are hard to obtain for most professionals 5-10 years out of school unless they've made a particular effort to stay in contact with undergrad faculty members. They understand this and take it into consideration.
Unless something has changed recently, the letters of recommendation are pretty much a formality. If you have a bachelor's in CS with decent grades, getting admitted isn't difficult. I was pretty flippant about the whole thing, applied one afternoon on a whim after reading about it on HN. Asked my manager for a letter of recommendation as well as my nearest colleague. No letters from anyone in academia.
I think the people who have the most difficulty getting accepted are those without a bachelor's in CS who also don't have some good CS fundamentals courses to show achievement and interest.
I did complete the program, and I am happy for the accomplishment. But with my experience (I started working in the mid 90s) this wasn't for my career, it was for my own satisfaction. But in addition to being glad for the achievement, I was soooo glad to be done, LOL. The real commitment is not financial, it is time.
I have not applied to this program but I've gone through a part-time MBA. I doubt that an online program for working professionals is as rigorous about references as, say, undergrad admission to an Ivy League program or Oxford or Hogwarts or something. Just get a couple of coworkers with a similar advanced degree to write something that says "this person exists and I think they can handle the load" and you'll be fine. Remember that college is a business; if you look like you can both handle the program and pay for it, they'll let you in.
I hit the same roadblock unfortunately. My academic references were all in a different field and I hadn't really stayed in contact except with one professor, who sadly has died. I did see that there's an option to use professional references, so even though I haven't done this myself, one route you could consider taking is to get references from managers, colleagues etc. who can speak to your technical knowledge. I agree though with your general point that after being out of an academic environment for a while that requirement becomes challenging.
I don't think OMSCS is that selective. Get a couple of letters from your former professors and/or bosses. Letters are supplementary, not the sole determinant. More than the letters, they likely care about your GPA and GRE.
I remember reading this when you published it! I was considering going at the time, as well as some years after, but this + work + starting a family had me choose different. I don't know if you've written a follow-up, but I would 100% read that.
In the past they made videos available via Udacity, which were removed after Udacity turned their focus to short & easy (which often means superficial) courses for enterprise training instead of "serious" university courses. I guess that was not a viable business.
Of course they did not come with any assignments, just like these courses. Can't blame them, but other universities offer much resources -- for the same topic, you can often find a course offered by another university that provides videos hosted on YouTube, full assignments and labs, even exams. The only thing you are missing is TA/office hours and the course credit. In other words, unless you actually want to earn credits and work towards a degree, I suggest that you skip OMSCS videos unless there is no alternative.
I was the head of enterprise curriculum in 2018 and an OMSCS grad in 2016. This was a weird time to work for Udacity and the company went thru a major shakeup in 2019. The “breakup” with GT happened before the focus on enterprise and the enterprise focus was somewhat short-lived as the CEO was replaced just as enterprise was ascending as the primary revenue stream. COVID was rough for Udacity, and content production was commoditized.
Udacity recently released their own MS AI though with shaky accreditation via Malta's MFHEA which was recently rejected by EQAR, making it unaccredited abroad.
Princeton used to offer algorithm courses on Coursera with full assignments including auto grading, I don't know if that's still the case.
I don't know enough about (or have time to learn) other offerings, but I am sure there are a few more out there. Class central should be a good place to discover those courses.
UTexas seems to be crushing it in the ML/AI space as they offer far more recent courses with deeper topics; for everything else OMSCS is probably a better choice even though it has a relentless pace of busywork making even easy classes draining. Stanford OTOH is like GT and UT merged together (both crazy difficult projects and a lot of math), but at 2x the pace. UT is way more relaxing than either, one can take 3 courses alongside a job and be fine, which is next to impossible at GT and Stanford. Conversely, if one wants to continue by doing research, Stanford and GT are much more useful due to ample opportunities to do so.
GT used to have a collaboration with Meta's FAIR and one could do final projects in the DL and NLP classes on recent research topics from Meta AI; some people ended up at top conferences with their work. Not sure how it's going to be now given Meta effectively disbanding FAIR in favor of Wang's Superintelligence. GT also has team research project under VIP (Vertically Integrated Projects) where a prof leads a team of students towards new findings; there are also a bunch of PhD seminars one can take. Stanford has a plenty of CS3xx research classes one can take where the final project has to be some novel research.
Readit News
The content is great, and most of it is available on Open Courseware, YT, etc, but here's what else you get by officially going through the program:
- the amazing community of TAs
- the assignments
- the feedback on reports & projects (either automated, or through TAs)
- the collaboration with other students on Ed, Discord, Slack, etc
- the forcing function of deadlines, having to study for exams, etc
- free access to academic libraries, IEEE, ACM, O'Reilly, etc
- access to software and services, educational packages from GitHub, Wolfram, Google Colab Pro, student discount in a bunch of places, etc
Another underrated aspect is GT's ability to preserve rigor of the program overall, despite the scale and number of students in some courses (the most popular ones have 1,000-1,500 students per semester).
If you're on the fence on applying, I strongly recommend you do. The program is affordable enough that there's no harm in trying for a few semesters to see if matches what you're looking for.
Glad to answer any questions.
My resume is also looked at differently after mentioning Georgia Tech. It really helped gain a lot of confidence. Fundamentally changed things as my undergraduate in India was not a good experience for me.
You may need to tweak for different courses, but I've used for ML4T, GIOS, and ML, and it has been incredibly helpful.
[1] https://github.com/guiambros/vidcat
I wish! I travel quite a bit for work, so it breaks my legs every time it happens. Plus family, kids activities, etc. ML was brutal this semester, but hoping the curve will help a bit.
But it's ok, slow and steady is the way to go. Besides, I'm doing this for the fun of it; I don't need the diploma for career or anything.
See you around!
Unfamiliar with us academic terms.
That course was great, though, and I definitely learned some things I'm glad to have learned!
IMO the instructional materials are a small part of the value. The things that stood out to me were:
- the assignments
- the autograding of programming assignments
- giving and receiving peer feedback about written assignments
- learning some LaTeX for those assignments
- having an artificial reason (course grade) to persist in improving my algorithm and code [on the problems taught in that course, I wouldn't have been self-motivated enough if they were just things I came across during a random weekend]
* My fellow classmates. Had a small study group where we got on Discord to hang out and it was a blast
* The TAs - they were so dedicated to the students and fantastic. MVPs of the program
I wasn't in any discord groups but the class discussion forum was a nice community.
I would not let the lack of assignments, tests, and quizzes stop you from trying these if you are interested. At a minimum, they would give you a feeling for what the program/s are like, and possibly encourage you to enroll into the online degree program, which is an exceptional value.
How do you all deal with this?
I think the people who have the most difficulty getting accepted are those without a bachelor's in CS who also don't have some good CS fundamentals courses to show achievement and interest.
I did complete the program, and I am happy for the accomplishment. But with my experience (I started working in the mid 90s) this wasn't for my career, it was for my own satisfaction. But in addition to being glad for the achievement, I was soooo glad to be done, LOL. The real commitment is not financial, it is time.
Of course they did not come with any assignments, just like these courses. Can't blame them, but other universities offer much resources -- for the same topic, you can often find a course offered by another university that provides videos hosted on YouTube, full assignments and labs, even exams. The only thing you are missing is TA/office hours and the course credit. In other words, unless you actually want to earn credits and work towards a degree, I suggest that you skip OMSCS videos unless there is no alternative.
Princeton used to offer algorithm courses on Coursera with full assignments including auto grading, I don't know if that's still the case.
I don't know enough about (or have time to learn) other offerings, but I am sure there are a few more out there. Class central should be a good place to discover those courses.