Adamah

Err, can't say I agree with that advice, Icy. First, there are very few reasons to turn down a PhD in CS at MIT for a MS in CS anywhere, especially if that MS doesn't come packaged with a RAship. Second, I disagree that it's "extremely easy" to get a RAship. Professors there tend to have little time for students, even grad students, and generally look down on MS students.

Do you have to commit to being a TA at Stanford for the whole Masters?
 
I visited Stanford when I was accepted to the MS program and I wasn't that impressed. Both the students and professors seemed to treat it as a 5th year of undergrad designed to give students a higher starting salary when going into industry. The emphasis isn't on research, though I got the impression it wasn't impossible to go onto research if you wanted to. I also got the impression that professors there look down on MS students, so it might be hard to find an advisor for research there.
 
I don't think you'd get much out of the Stanford MS. If you're a TA you wouldn't have much time for research, and I don't think you can count on finding an advisor anyway. That said, if you do get two upper-tier publications this semester, then that would work in your favor when applying a second time (did you mention that you had these publications underway when you applied the first time?)
 
Cornell is probably your best bet. It's one of the best PhD programs out there anyway (assuming Information Science isn't that different from their CS program). At that level, your career in academia will be determined by your research, not which school you went to.

I had the same choice last year when I applied to PhD programs. I chose CMU. A few things to consider:
 
1) CS rankings do matter, especially in academia (and obviously, it's the program ranking that matters). A professor at U Mich told me that when I asked him about their hiring process for new faculty. When it comes to reputation, there is a big gap between the "Top 4" and other universities. The Harvard name will help you pick up girls at a bar. The CMU name will gain you respect among your peers. (A couple caveats to this: First, I don't think it should be that way in either direction, but that is the reality we live in. Second, Harvard does still have a good name in CS, and CMU does still have a good general reputation in the US.)
 
2) Pittsburgh is a much better city than you seem to think. I also thought that Pittsburgh would not be a great place to live (and I was comparing this against Stanford), but I've really enjoyed it here. It has a great academic community, a lot of culture, the people are friendly, and the city is very safe. Plus, living expenses are about half price compared to Boston, so you can live like a king here on your stipend. I would complain about winter, but you'd get way more of that in Cambridge.
 
3) The people at CMU, both the students and the faculty, are simply great people. When I visited Stanford and asked students why I should go there, they said "because it's Stanford!" When I visited U Mich, nearly every presentation had a reference to the university being top-10 in X many subjects. When I visited CMU, the university's ranking was hardly mentioned. At one point I even asked a student if CMU Robotics is indeed the top program in the country, to which she replied "that's what I've heard." In short, there is no pretentiousness at CMU.
 
4) Advisors are a very important factor. You should make a list of faculty at CMU that you'd like to speak to and send it to the coordinator. I'm sure they'll schedule Skype talks with them.

Sorry for reviving an old thread, but I don't check this site very often anymore and every once in a while I get messages from people in a similar situation asking how my choice turned out. I thought I'd just post an update here.
I'm very happy with having decided to go to CMU. I ended up getting a RAship after the first semester (full funding + stipend). The Robotics MS program is indeed fantastic -- most people get funding (it varies from year to year, but my year ~90% received funding) and the expectations and social interactions are essentially identical to the first 2 years of the PhD program. People quickly forget who is a Masters student and who is a PhD student. This is not the case at almost any other MS program in other universities or even at CMU (I think LTI at CMU is another excellent example of this though, from what I've heard). I eventually continued to a PhD program at CMU and am very happy there.
Looking back, the decision to go to CMU should have been a no-brainer in my situation. When I visited Stanford it was clear to me that they don't really care about their MS students, and when I visited CMU it was clear that they did. At least on the CMU side, my impression proved accurate.

I think CMU MS Robotics is the better choice. You should contact the administrators you're in touch with at each school and ask them what percentage of students landed an RAship. Doing an RAship is essential if you want to pursue a PhD at a top program. The CMU MS Robotics will be 100% because it is a program requirement, with probably ~90% of the students receiving funding. Stanford's rate will be lower. You could try talking to faculty at Stanford now and seeing if any of them would be interested in taking you as an RA if you go there, but you will probably not get any firm commitments.
My statements about CMU treating their MS Robotics students the same as PhD students is still true. As an example, there are two different grad-level intro. to machine learning courses as CMU: 10-601 and 10-701. 10-601 is the more practical version, intended mostly for MS students. 10-701 is the PhD-level course for those interested in ML. The MS Robotics counts only 10-701 toward your course requirements (or you can take an AI course instead of the ML course).
Regarding the cold weather, that's a personal preference, but I personally don't think you should base your career on which school has better weather.

Sorry for reviving an old thread, but I don't check this site very often anymore and every once in a while I get messages from people in a similar situation asking how my choice turned out. I thought I'd just post an update here.
I'm very happy with having decided to go to CMU. I ended up getting a RAship after the first semester (full funding + stipend). The Robotics MS program is indeed fantastic -- most people get funding (it varies from year to year, but my year ~90% received funding) and the expectations and social interactions are essentially identical to the first 2 years of the PhD program. People quickly forget who is a Masters student and who is a PhD student. This is not the case at almost any other MS program in other universities or even at CMU (I think LTI at CMU is another excellent example of this though, from what I've heard). I eventually continued to a PhD program at CMU and am very happy there.
Looking back, the decision to go to CMU should have been a no-brainer in my situation. When I visited Stanford it was clear to me that they don't really care about their MS students, and when I visited CMU it was clear that they did. At least on the CMU side, my impression proved accurate.

With a 2.64 GPA, I don't think you'd get into any PhD programs. If you think you have since changed and the 2.64 GPA is not an accurate reflection of your abilities, then you'd probably be best off going for a MS first. This would allow you to show a better GPA to PhD programs (ideally a 4.0) and argue convincingly that the 2.64 GPA is not an accurate signal of your abilities. It would also potentially allow you to do research during your MS.
However, keep in mind that the MS has a high cost, especially if you end up not going for a PhD. You have to decide if that's worthwhile.

Err, can't say I agree with that advice, Icy. First, there are very few reasons to turn down a PhD in CS at MIT for a MS in CS anywhere, especially if that MS doesn't come packaged with a RAship. Second, I disagree that it's "extremely easy" to get a RAship. Professors there tend to have little time for students, even grad students, and generally look down on MS students.

Do you have to commit to being a TA at Stanford for the whole Masters?
 
I visited Stanford when I was accepted to the MS program and I wasn't that impressed. Both the students and professors seemed to treat it as a 5th year of undergrad designed to give students a higher starting salary when going into industry. The emphasis isn't on research, though I got the impression it wasn't impossible to go onto research if you wanted to. I also got the impression that professors there look down on MS students, so it might be hard to find an advisor for research there.
 
I don't think you'd get much out of the Stanford MS. If you're a TA you wouldn't have much time for research, and I don't think you can count on finding an advisor anyway. That said, if you do get two upper-tier publications this semester, then that would work in your favor when applying a second time (did you mention that you had these publications underway when you applied the first time?)
 
Cornell is probably your best bet. It's one of the best PhD programs out there anyway (assuming Information Science isn't that different from their CS program). At that level, your career in academia will be determined by your research, not which school you went to.

I was in a similar situation 3 years ago. I had been accepted to U Toronto and UMass for the PhD programs, and Stanford and CMU (Robotics) for the MS programs. I ultimately chose the CMU MS Robotics, and am now in the CS PhD program at CMU. However, my outcome is not necessarily typical and there are a lot of factors to consider.
 
First of all, the CMU MS in Robotics is a very research-focused Masters. Most of the students get funded RAships (in my year, almost all did, though in some years it is as low as about half). All students spend a lot of time working on a Masters thesis, and the requirements for the first 2 years are identical for the PhD and MS students. This allowed me to spend a lot of time on research while in grad school and get a top-tier publication.
 
Like you said, the MSCS and MSCV programs at CMU are not research based, so if your goal is a PhD then you should avoid them. I don't know much about the MSML program (I believe it's fairly new), but it might be a good option. I would ask them how many students get funded RAships each year, and where past students have gone afterwards.
 
I would also contact professors you are interested in and see if there is any potential for an RAship with them. I think many will be open to taking on a MS student for research if they have funding. Do not expect any commitments, because the professors likely can't predict their funding situation that far out and they also will prioritize PhD students in the fall, but if you find a number of professors that are options then that's a good sign.
 
Going for the MS is definitely the riskier option. You might end up wasting 1-2 years on coursework that you'll have to repeat at another university, not to mention the money you'll spend. It really depends on if you think you can land a funded RAship at CMU and do a fantastic enough job that your advisor will push for you to attend full-time. Unless you get a really good vibe from some professors at CMU, I would lean toward Wisconsin or Michigan, both of which are good schools and may have many good advisors that fit your interests.

I was in a similar situation 3 years ago. I had been accepted to U Toronto and UMass for the PhD programs, and Stanford and CMU (Robotics) for the MS programs. I ultimately chose the CMU MS Robotics, and am now in the CS PhD program at CMU. However, my outcome is not necessarily typical and there are a lot of factors to consider.
 
First of all, the CMU MS in Robotics is a very research-focused Masters. Most of the students get funded RAships (in my year, almost all did, though in some years it is as low as about half). All students spend a lot of time working on a Masters thesis, and the requirements for the first 2 years are identical for the PhD and MS students. This allowed me to spend a lot of time on research while in grad school and get a top-tier publication.
 
Like you said, the MSCS and MSCV programs at CMU are not research based, so if your goal is a PhD then you should avoid them. I don't know much about the MSML program (I believe it's fairly new), but it might be a good option. I would ask them how many students get funded RAships each year, and where past students have gone afterwards.
 
I would also contact professors you are interested in and see if there is any potential for an RAship with them. I think many will be open to taking on a MS student for research if they have funding. Do not expect any commitments, because the professors likely can't predict their funding situation that far out and they also will prioritize PhD students in the fall, but if you find a number of professors that are options then that's a good sign.
 
Going for the MS is definitely the riskier option. You might end up wasting 1-2 years on coursework that you'll have to repeat at another university, not to mention the money you'll spend. It really depends on if you think you can land a funded RAship at CMU and do a fantastic enough job that your advisor will push for you to attend full-time. Unless you get a really good vibe from some professors at CMU, I would lean toward Wisconsin or Michigan, both of which are good schools and may have many good advisors that fit your interests.

I had the same choice last year when I applied to PhD programs. I chose CMU. A few things to consider:
 
1) CS rankings do matter, especially in academia (and obviously, it's the program ranking that matters). A professor at U Mich told me that when I asked him about their hiring process for new faculty. When it comes to reputation, there is a big gap between the "Top 4" and other universities. The Harvard name will help you pick up girls at a bar. The CMU name will gain you respect among your peers. (A couple caveats to this: First, I don't think it should be that way in either direction, but that is the reality we live in. Second, Harvard does still have a good name in CS, and CMU does still have a good general reputation in the US.)
 
2) Pittsburgh is a much better city than you seem to think. I also thought that Pittsburgh would not be a great place to live (and I was comparing this against Stanford), but I've really enjoyed it here. It has a great academic community, a lot of culture, the people are friendly, and the city is very safe. Plus, living expenses are about half price compared to Boston, so you can live like a king here on your stipend. I would complain about winter, but you'd get way more of that in Cambridge.
 
3) The people at CMU, both the students and the faculty, are simply great people. When I visited Stanford and asked students why I should go there, they said "because it's Stanford!" When I visited U Mich, nearly every presentation had a reference to the university being top-10 in X many subjects. When I visited CMU, the university's ranking was hardly mentioned. At one point I even asked a student if CMU Robotics is indeed the top program in the country, to which she replied "that's what I've heard." In short, there is no pretentiousness at CMU.
 
4) Advisors are a very important factor. You should make a list of faculty at CMU that you'd like to speak to and send it to the coordinator. I'm sure they'll schedule Skype talks with them.

I don't think you need more CS courses, but I'm not 100% on that. Machine learning isn't core CS anyway. It's really more statistics, so I don't think the professors would care whether you know your data structures if it's clear your interest is in machine learning.

I don't know if CMU counts as Northeast, but you should consider it if you're interested in ML. They have a distinct PhD for ML, and they even have a subprogram for Neural Basis of Cognition. You would also definitely want to apply to MIT, Cornell, and UMass (which has one of the best Machine Learning programs in the country).

Your situation is pretty unique. You should definitely talk to some faculty. I'm sure they'd be happy to speak with you.

Agreed. Also, UPenn has very good placement for its grad students. I know of at least a few professors at top schools that came out of UPenn.

I agree with Oh Yeah. The reason a lot of students come from top undergrads has less to do with the school's name and more with who they can get letters from. I also don't think you'll get into Berkeley, but it may still be worth it. In the end, $80 isn't that much.