Princeton vs. Columbia, CS PhD

[cabby]

My interests are machine learning and vision/graphics. Which of Columbia or Princeton would you choose?

Princeton maybe has a better name and a higher CS ranking, but Columbia has some really strong research and professors, although I'm not sure 100% what prof I'd end up working with there yet (although it seems flexible).

I like the research and professors at both places, but was wondering if one might offer better opportunities afterwards, particularly since I may want to work in academia.

Thanks!

[adelashk]

I would go with Princeton (better in engineering and CS, brand name, better job prospects).. I can assume the professors are more known at Princeton's CS dept.

[OH YEAH]

Princeton's ML department is pretty small, but high quality: Blei just won an ONR young investigator award and appears to be one of the rising stars in the field -- I don't know of many junior faculty with a paper cited 2700 times. Of course, Schapire is one of the most cited CS researchers in any field (I'm taking AI with him this fall, I hear he is a great teacher too). I don't know much about Yael Niv.

We don't have anyone doing computer vision from what I know, but the graphics group here is pretty solid -- Funkhouser, Rusinkiewicz and Finkelstein are all heavily cited.

I love it here. If you have any specific questions, I'd be happy to answer them!

Edited April 12, 2011 by OH YEAH

[cabby]

We don't have anyone doing computer vision from what I know, but the graphics group here is pretty solid -- Funkhouser, Rusinkiewicz and Finkelstein are all heavily cited.

I guess one of the main differences is that Columbia has a very strong vision group, and Princeton doesn't currently have a dedicated vision person, but Funkhouser and Rusinkiewicz do work on vision projects to a lesser extent.

I definitely have an interest in vision which makes me think Columbia, but I guess Princeton might be just as good, except I'd more be working in graphics/vision. Also, vision isn't my only interest, so I'm worried about basing the whole decision on this. I really am interested in many projects going on at both places but was wondering if one would offer better opportunities, especially for academic positions afterwards.

Does anyone else have an opinion/advice?

[edvolkov]

I guess one of the main differences is that Columbia has a very strong vision group, and Princeton doesn't currently have a dedicated vision person, but Funkhouser and Rusinkiewicz do work on vision projects to a lesser extent.

I definitely have an interest in vision which makes me think Columbia, but I guess Princeton might be just as good, except I'd more be working in graphics/vision. Also, vision isn't my only interest, so I'm worried about basing the whole decision on this. I really am interested in many projects going on at both places but was wondering if one would offer better opportunities, especially for academic positions afterwards.

Does anyone else have an opinion/advice?

Funkhouser and Rusinkiewicz have more citations in last 10 years altogether than three Columbia profs working in vision.

[cabby]

Funkhouser and Rusinkiewicz have more citations in last 10 years altogether than three Columbia profs working in vision.

Just curious, which were the three profs from Columbia you compared them to? Also, what's a quick way to count citations?

Thanks!

[edvolkov]

Eitan Grinspun, Ravi Ramamoorthi, Steve Feiner (I don't know how this set is representative). You can use "publish or perish" program to count citations.Don't use microsoft academic research to do it since it's very buggy up to this moment.

[explorer-c]

Eitan Grinspun, Ravi Ramamoorthi, Steve Feiner (I don't know how this set is representative).

No. That is not a representative set. For one, it doesn't include Shree Nayar, arguably the most famous vision guy in Columbia. Anyway, if you have heavy interest in computer vision, then there is no contest, Columbia is much better. If you're leaning more towards ML, then Princeton is better. Just so you know, last year I was deciding between Columbia, Princeton and a bunch of other places. But since I have a pretty specific interest in computer vision, I ended up not going to either of those places, but Columbia was on my shortlist.