I've been wondering for a while whether it's worth enrolling in [the OMS CS program](https://www.omscs.gatech.edu/program-info/specializations) and after thinking for some weeks I came to the conclusion that it's a good opportunity to take. I decided that I'll try to enrol in the next possible semester.<p>From the four specializations they offer I'm only interested in Computing Systems or Machine Learning. I believe Computing Systems is the way to go but Machine Learning seems pretty nice with a good future in innovation meanwhile Computing Systems looks like too theoretical - not in a bad way.<p>To put it in other words, Computing Systems sounds like academic oriented and Machine Learning looks more oriented to technology development. So I'm asking for your opinions or experiences if you have already taken these courses. Were they good? Do you recommend them?<p>Also, more in specific I'm having problems understanding how I'm supposed to chose the courses I want to take. I'm not familiar with the US college system.<p>I understand I have to take 30 hours (10 courses), which are divided in core and elective. However, for the [Specialization in Computing Systems](https://www.omscs.gatech.edu/specialization-computing-systems) it says it's 18 hours, 3 core courses and 3 elective and for [Machine Learning](https://www.omscs.gatech.edu/specialization-machine-learning) are 15 hours, 2 core courses and 3 elective. Where are the other remaining hours?<p>Some background, I'm Mexican with a bachelor degree in Computer Systems Engineering.
I completed my OMSCS last Spring, specializing in Machine Learning. My advice to you would be to not worry about any specialization as you can change it later on. Focus on taking courses that interest you. You need to complete 10 courses for completing the program. With 2-3 mandatory courses + 3 electives, you have room for at least 4 other courses not in your specialization.<p>There are some interesting courses on the Computing System track. FWIW, my undergrad was in a different engineering field. So I really enjoyed the Network Security and Intro to Operating Systems. The Algorithms course reshaped my perspective on formal methods and correctness. The HCI course is also very good.<p>As mentioned by @bewestphal , check out <a href="https://omscentral.com/" rel="nofollow">https://omscentral.com/</a> for reviews to get an idea about the popularity, difficulty and time investment for the courses.
I'm in it right now, starting my third term in May. It's been very valuable for me. However, I think I'm like most of the students: coming from a math, engineering, or science background, transitioning into a more software development-oriented career. The value of the OMSCS degree is very clear for us: the degree as a qualification, and exposure to knowledge/skills we don't already have.I'm not sure if it would be as valuable for someone who already has some sort of computer science/systems background. I'm not really sure what a computer systems engineering degree entails so I may be way off here.<p>Some other thoughts:<p>The time investment and effort to do the courses is serious. Not only are the courses very time-consuming (10 - 30 hours per week), but you're also expected to do research and study up on material on your own in certain classes. You can't just be a passive consumer of the material. It can be very rewarding and/or very painful depending on what you're getting out of the course.<p>Your choices are constrained by class capacity. It is very difficult to get into the more popular courses (i.e the ML ones in particular) and so be prepared to take non-ML courses until you have enough seniority to get better registration preference. Don't even bother trying to plan out your courses in advance, a lot of it will just be determined by when you are permitted to register for classes! To be fair, based on recent emails from GT, it seems like they're aware of this problem and looking to open up other classes (possibly taking advantage of more courses going online during the pandemic). They also seem to be opening up courses that were formerly restricted to the OMSA degree (the more math-heavy analytics equivalent) which I'm really excited about.
UT Austin's program is also worth considering. It's newer but there is supposedly more consistent quality and rigor to the courses. Of course nobody has taken both so it is speculation from students at both schools, but that is the general consensus I could pick up from the respective subreddits for both programs. You also have to take the GRE for UT Austin.<p>As for your question about remaining hours, they are classes you choose. Anything not explicitly required or part of your specialization gets filled by electives of your choosing. You can also see rankings by students of all the courses here: <a href="https://omscentral.com/" rel="nofollow">https://omscentral.com/</a>
The remaining units are free electives if your choice. More than enough to touch all topics you’re interested in.<p>Definitely check out the reviews here
<a href="https://omscentral.com/" rel="nofollow">https://omscentral.com/</a><p>Personally I recommend Computer Vision, Reinforcement Learning and AI Techniques for Robotics. Skip classes that emphasize writing essays over coding.
Anyone complete or accepted to OMSCS without a background in Computer Science?<p>Any recommended ways to fulfill their requirements for the basic CS without paying $$$ to take courses?<p>Was denied for not having the CS background before joining a FAANG.
May want to try GradCafe for nuanced responses. Not sure if you're talking about MS or MEng, but if it's MS then finding a good match with your advisor and research group is more important than anything else.