MS in Stanford CS v/s CMU Robotics

[meditatingmuse]

I have to decide between the MS program in CS at Stanford and MS in Robotics at CMU.

I don't know yet whether I want to continue for PhD but I do know that I don't want a programming job after graduation. I have some research background in machine learning and computer vision, and am hoping to do research in grad school as well.

I think I have narrowed down the decision to deciding between the following pros of each program, and I am eager to hear your thoughts on which of them have more weight.

The main pro of CMU in my case is that there is greater likelihood of getting research experience, and that too in a field of interest (as there are more AI/ML/Robotics faculty and relatively fewer admits) compared to Stanford. Most current students from CMU that I have spoken with got RA positions from the 2nd semester of the program. And the pros of Stanford are the exposure to entrepreneurship and the benefits of a better overall ranked university.

Looking forward to hear your suggestions, thanks!

 

[rising_star]

Do you have any funding? The cost of living in Pittsburgh is much lower than that in Palo Alto, so that's one thing to consider. Also, if you don't want a programming job, why do a degree in computer science over one in robotics?

[meditatingmuse]

Neither programs ensure guaranteed funding. Although on scanning the student list of the websites of both programs, it seems that about 5% of Stanford MS students have RA ship. I am not sure about TA ship, but a current student mentioned that many (if not most) students are able to TA their way through the program. On the other hand, most current students that I spoke with at CMU RI have funding from 2nd semester as an RA. This is not surprising, since the ratio of total admits to total faculty seems to be much lower at CMU. But to be honest, while funding is important, I am not looking at it as a make-or-break criterion.

Stanford's MS has a research track and also specialization options in AI, HCI, Software Engg etc. If I join Stanford, I hope to do research in AI, which would hopefully prepare me for non-prorgramming jobs as well (or a PhD).

 

 

 

[svent]

I hear Stanford's MS class is huge (twice as big as PhD) so it's hard to get an advisor to supervise a thesis (especially since they care more about PhD students), and I imagine TAships are very competitive at the Master's level.

[blacknighterrant]

On 3/20/2016 at 3:51 PM, meditatingmuse said:

I have to decide between the MS program in CS at Stanford and MS in Robotics at CMU.

I don't know yet whether I want to continue for PhD but I do know that I don't want a programming job after graduation. I have some research background in machine learning and computer vision, and am hoping to do research in grad school as well.

I think I have narrowed down the decision to deciding between the following pros of each program, and I am eager to hear your thoughts on which of them have more weight.

The main pro of CMU in my case is that there is greater likelihood of getting research experience, and that too in a field of interest (as there are more AI/ML/Robotics faculty and relatively fewer admits) compared to Stanford. Most current students from CMU that I have spoken with got RA positions from the 2nd semester of the program. And the pros of Stanford are the exposure to entrepreneurship and the benefits of a better overall ranked university.

Looking forward to hear your suggestions, thanks!

 

CMU is by far the better option. CS degrees are not particularly useful outside of cs, so it would be harder to get an engineering job with one than a robotics job. In addition you say you want to continue to PhD and research is extremely important when applying to PhD programs, if CMU is where you can do research then you need to go to CMU.

[meditatingmuse]

@svent Yes, that seems true. From my experience at CMU so far (I am here for a research internship), it seems the program is more invested in MS students than Stanford.

[meditatingmuse]

@blacknighterrant

That sounds reasonable. However, most of my experience in robotics so far has been on the software side (at a robotics club in my undergrad and during my current research internship at CMU). I am concerned whether 2 years is "enough" bootstrap time to get a job in robotics after graduation (if I decide not to go for PhD)? Or whether PhD would become a requirement if I want a job in Robotics. Any thoughts?

[ender wiggin]

I'm deciding between Robotics PhD at CMU and ME PhD at Stanford so hopefully I can offer some useful advice... CMU definitely has smaller number of people and would be a bit easier to get with a specific professor but not much. The number of professors in AI/ML at CMU is greater. Both schools have amazing quality research so I wouldn't let that deter you. If you're deciding between PhD and getting a job, Stanford professors have ridiculous connections with silicon valley so that would be a great in for you to have. I think you'd easily get a robotics job with an MS. Also from my personal opinion the life at Stanford seems much more relaxed and exploratory than CMU which appears quite intense, research = life, and immersive so either can be good depending on what you want.

[blacknighterrant]

13 hours ago, meditatingmuse said:

@blacknighterrant

That sounds reasonable. However, most of my experience in robotics so far has been on the software side (at a robotics club in my undergrad and during my current research internship at CMU). I am concerned whether 2 years is "enough" bootstrap time to get a job in robotics after graduation (if I decide not to go for PhD)? Or whether PhD would become a requirement if I want a job in Robotics. Any thoughts?

If you think about it undergrad may be 4 years, but over half that time is usually used for none core information (general ed requirements and electives), many schools don't even require students to choose a major before the end of their second year, so only 2 years in a bachelors is really important for career building. 2 years is more than enough time for getting the information you need and you'll be at an advantage when applying for jobs because you have a masters anyway, just make sure you use your time productively (co-ops in the summer and research during the year, the research and co-ops will give you a better understanding of what you really want to do.). Honestly although it is sad to say, the point of going to school at any level bellow PhD is not to learn or gain information, but to demonstrate that you can learn and function well to employers through validation in the form of a degree from the school. Most of what you actually need on the job you will learn while working or teach yourself afterwords. Honestly, I don't think going robotics or cs was the best option, going for a robotics degree is also kind of limiting, I have the same general interest but went for electrical engineering since it is more general and would allow you to apply for robotics jobs in addition to other engineering jobs more easily. In terms of the PhD being a requirement for a robotics job, it will depend on the job, but a PhD being required is very very unlikely, the majority of people working probably have a bachelors, while a few others have masters, and very few people have PhDs. PhDs are more for those that want to remain in academia than those that wish to work in the industry, the main consideration is that they are much too specific (you research an extremely specific subtopic in robotics), and normally the specific knowledge you get in your PhD studies is not worth the extra cost companies would have to pay to employ you. If your sure that working in the industry is what you want to do, my best advice for you would be to look at job postings you would be interested prior to applying for PhD programs to see what they are looking for and how many of the listings require a PhD, asking for advice during your co-ops would also prove to be useful.

[meditatingmuse]

I guess the biggest benefit of Stanford the way I see it is the entrepreneurship culture. Does the exposure to VC funding and similarly minded peers seem to help? Although I have heard that Stanford MSCS is seen as a "cash cow", which might undermine the effect of this exposure.