CompSci PhD
November 7, 2004 10:00 AM   Subscribe

I'm looking for a good computer science Ph.D. program to apply to, with strong research groups in both bioinformatics and AI.

I'm graduating from a leading US university with a triple degree in CS, applied math, and statistics. My research background is fairly strong (a few joint published papers in my research concentrations, recommendations, etc.), my GREs are/should be very decent, but what's holding me back is my GPA, which really sucks except for the past two semesters. Because of this, the top-tier schools (Stanford, MIT, Princeton, Berkeley, probably CMU) are out.
Therefore I'm looking mainly for schools just below that level. I can apply as either CS or Applied Math (I can probably collaborate with the right people from either position), but I'd rather get into CS. My requirements are:
  • Strong research groups in both of my research concentrations - bioinformatics (comparative genomics) and AI (machine vision, computational cogsci).
  • Decent location (I don't want to be stuck in Idaho or the rural Eastern US, for instance. I'd rather visit them on hiking trips.) A large northern coastal metro area would be ideal.
Can't think of anything else right now...
I will consider Canadian, UK, and other English-speaking schools, but I'm told they don't have nearly as much money to assist students financially as the US schools. Is that correct?
I have a list of about 40 schools, but that's too many (I'd like to narrow it down to at most 25) and I don't know how comprehensive my initial list was.
I'm told the proper way to select schools in this situation is to read the current publications and identify which places the strong research groups come from. I'm in the process of doing that, but wanted to ask for more advice regardless.
posted by azazello to Education (17 answers total) 1 user marked this as a favorite
 
edinburgh?
i know it use to be ok with ai and that there are local jobs in bio-informatics...
posted by andrew cooke at 10:59 AM on November 7, 2004


Rutgers? Penn? Brown? Irvine? My husband works in computer/human interaction research and those schools are where some of his lab's recent hires got their Ph. D.s.

I have to say that, if I were you, I would bite the bullet, take out student loans, and go somewhere outside the US and get a Master's (perhaps the program in the Netherlands that Andrew Tanenbaum is running?), do well, and then come back and apply to top-echelon Ph. D. programs as well as for NSF support. If you do well in an overseas Master's program, your poor college GPA won't be such a factor in your Ph. D. applications.
posted by Sidhedevil at 11:19 AM on November 7, 2004


I believe that UNC-Chapel Hill is supposed to have the nation's top bioinformatics program (or at least it did a few years ago), and its CS program is tops too (David Brooks, author of The Mythical Man Month is emeritus there). I've never heard UNC associated specifically with AI though.
posted by gsteff at 1:17 PM on November 7, 2004


UMass Amherst might be interesting to you - they have a very strong AI program and growing bioinformatics. It is not in a city, but it is about 2 hrs from boston, 4-5 from nyc.

Disclosure: I went there as an undergrad, and worked in one of the AI research labs.

Also, you may find that GPA matters less than you think with strong recommendations from well-known people.
posted by advil at 1:17 PM on November 7, 2004


I am a graduate student at one of those top-tier schools, though not in CS, and can strongly second what advil is saying, with personal experience and that of my friends.

The fact that you have done research, have published research, and know what you want to do, puts you leagues ahead of smiling fresh faces with 4.0 GPA.

It sounds like you have a long list already, but don't hurry to cross places off the list on GPA alone. The departments out here put it third of the four numerical parts of the application, with only GRE being less important (the most important consideration, they say, is what you want to research, the second is recommendations).

Good luck!
posted by whatzit at 2:44 PM on November 7, 2004


I've heard decent things about U. Illinois at Champaign Urbana.
posted by Krrrlson at 3:48 PM on November 7, 2004


I'm at Michigan right now - it's a great department. But it doesn't have much in the way of bioinformatics, as far as I know, and I don't recommend the area.

Also, it was probably easier to get in when I applied, but my GPA was far from perfect.
posted by transona5 at 4:19 PM on November 7, 2004


Krrrlson is right, UIUC's CS department has a great reputation, and I have many friends in the graduate program. However, I'm not sure how well it fits the "decent location" criterion. Central Illinois is pretty much one gigantic farm, and you're two hours from Chicago.

Another tip I heard: apply for the spring semester. They demand for spots supposedly doesn't outstrip supply quite as much as it does during the Fall.
posted by Eamon at 4:53 PM on November 7, 2004


andrew cooke is right - Edinburgh would be ideal for you, assuming that the AI department is as good as it was fifteen years ago. Back then I know that they had a lot of US folks doing Ph.D's.
posted by thatwhichfalls at 6:57 PM on November 7, 2004


think about mcgill university, in montreal. (note: yes, i am biased.) they have a strong AI group and a strong Bioinformatics group, both of which I have some familiarity with. I'm doing my phd in discrete math (also a strong group) but my masters was in bioinformatics/computational biology. the bioinfo people and the ai people like to work with each other, too, which is probably a plus from your perspective. the academic climate in CS at mcgill is very open and non-heirarchical, which i value greatly.

you also have the benefit of living in montreal, the best city in the world :P.
posted by louigi at 7:32 PM on November 7, 2004


Is McGill French-language instruction or English-language instruction? (And I love Montreal. A lot.)
posted by Sidhedevil at 8:29 PM on November 7, 2004


If you're considering studying in Canada I wouldn't worry about financial support; most departments (at least in the sciences) will not admit you unless you are financially supported (either by a fellowship from an external agency or directly from your supervisor’s grant).

While I can't say much about the bioinformatics programme at McGill specifically, I can tell you that McGill has an excellent international reputation, especially in the biomedical sciences, so you might want to look into their programme (Note: I'm also a Ph.D. student at McGill so my advice is perhaps a bit biased as well).
posted by trident at 9:55 PM on November 7, 2004


The language of instruction at McGill is English, but francophone students are permitted to write papers and exams in French.
posted by trident at 9:58 PM on November 7, 2004


While Canadian graduate students in the humanities and social sciences are not well funded (some are entirely without, many have partial), I believe that this is not so much true in the case of sciences, and computer science would be up there for funding (there is stil a demand for CS Ph.D. students). It's probably at least partly due to the different funding model (i.e. being funding from an advisor's grant).

When applying to graduate school, I was told that while GPA mattered, it was just one factor of many - also the GRE, reccomendations, writing sample, etc. Some time off or a masters would certainly make your GPA matter less. The publications definately help.

I know this is very hard (I had no idea where to apply), but are there certain researchers whose work inspires you? You could scan the faculty lists of the programs, and pick out those whose work you know or whose work sounds very interesting. This could help narrow your selection of schools. And you can always write to them asking if they are accepting advisees; my current advisor was very friendly when I did so, and even gave me tips on applying successfully.
posted by jb at 11:49 PM on November 7, 2004


Queen's University in Kingston, Ontario has the only undergraduate program in biomedical computing in North America (if I remember correctly), and has been hiring bioinformatics profs like mad -- thus making lots of opportunities for grad students in the area. Great school, did my MSc and BScH there. Contact me if you need more info.
posted by krunk at 9:35 AM on November 8, 2004


UMass Amherst might be interesting to you - they have a very strong AI program and growing bioinformatics. It is not in a city, but it is about 2 hrs from boston, 4-5 from nyc.

Disclosure: I went there as an undergrad, and worked in one of the AI research labs.

Also, you may find that GPA matters less than you think with strong recommendations from well-known people.
posted by advil at 1:17 PM PST on November 7


Seconded. Just recently graduated with a CS degree from UMass and while I regrettably did not have the time to get involved with research (long story) I still kept my eyes open. There's a lot of robotics and AI work going on, and yes, I believe there's at least one or two faculty members working on bioinformatics as well.

The school is well known for its strong post-graduate CS program (in fact we CS undergrads often resented that fact, heh) and while it's not in a major city, as advil pointed out, the Amherst/Northampton area still has plenty to do when you're not constructing B-trees or programming agents. Also as advil said, Boston is a 2.5 hour drive east, and NYC is a 4 hour drive south/southwest.

Plus, there are 4 other internationally known schools around, and while I don't believe any of them have nearly as good a post-grad CS program, there's still a lot of knowledge floating around the valley that can be taken advantage of.
posted by cyrusdogstar at 11:44 AM on November 8, 2004


Response by poster: Thanks everyone for your responses! I'll include the schools you mentioned in my search (almost all of them were already on the list but they probably won't be the ones kicked off of it).
posted by azazello at 10:34 AM on November 9, 2004


« Older Moral Calculus.   |   What can I do, ergonomically speaking, to work... Newer »
This thread is closed to new comments.