Applying to PhD in EECS -- Chances and comments?

Graduate School
1

<p>Hi guys, I've been lurking in this forum since last year. I read many people post their stats for others to comment on it but I can't seem to find people with stats similar to mine. Curiosity got the best of me so here I am posting my stats, hoping that you guys can comment on it with respect to my selection of schools. :)</p>

<p>I'm planning to apply to EECS, my area is computer vision and these are the schools that I'm interested in: Stanford, MIT, Berkeley, CMU, Washington, UIUC, Princeton and probably 1-2 other schools.</p>

<p>Undergrad university: Well known university outside US. I graduated in 2008.</p>

<p>Overall GPA/Major GPA: 3.3/3.5
This is probably the weakest aspect of my application :( but I'm hoping that my other stats will balance it out. I did excellent in my 1st and 4th year but bombed the rest because I'm not interested in the subjects (I took Computer Engineering as a major not knowing that I don't have that much interest in hardware, which is the main topic in my 2nd and 3rd year).</p>

<p>GRE: 800q, 450v, 3.5W
Basically, I took a GRE without preparation to see first how bad I'd do and go from there. As expected, my verbal and AW score is horrible :(. But I'm confident I can raise my v to >500 and my AW to >4.5. I'm planning to take another GRE but do you guys think it's necessary?</p>

<p>Research experience:
- Student research assistant in my 3rd year. During that period, I worked with a graduate student to come up with new algorithm for clustering in data stream. Work published in DASFAA 2009 (good conference in data mining).
- Paper from my final year thesis will be published in ICCV 2009 (top conference in computer vision) and I'm the first author of that paper.
- From my graduation (2008) until now, I work as a research assistant with a well known professor in my undergraduate university. The research area is computer vision and we're planning for a publication in CVPR 2010 but I'm probably not the first author of that paper.</p>

<p>Publications:
- DASFAA 2009 (good conference in data mining): 4th author
- ICCV 2009 (top conference in computer vision): 1st author
- Will submit a paper to IEEE Transactions of Medical Imaging (top conference in image processing area): 1st author
- Will submit a paper to CVPR 2010: n-th author (basically I don't know where my name will appear in there :P)</p>

<p>Recommendations:
I will have 3 decent letters from professors in my undergraduate university. One is a fairly young professor, the other two are quite famous in their field. They're alumni of Princeton, Cambridge, Umich.</p>

<p>So that's my stats. I know that I'm probably on the low end of spectrum for GPA, but I don't think there are many applicants that have research credentials (especially publications) as good as mine. So how do you guys think I'd fare with other applicants in the schools that I'm interested in?</p>

<p>Your comments will be very appreciated! Thank you. :)</p>

2

<p>What is your undergrad institution? That can be a large factor in addressing “low” gpa’s. </p>

<p>Your LOR’s and research look excellent, well above the norm. Your gre’s are okay, but I would consider retaking them for the higher V and AW - your targets are good ones to aim for. Have you researched your intended field and matched it to professors at those universities? If they do not match your research interests they will not take you regardless of credentials.</p>

<p>I would not think the schools you indicated are necessarily out of reach, but your gpa may cause you some problems. MIT and Stanford in particular are apt to attract students with your research experience who ALSO have the high gpa. Worth a try, though.</p>

3

<p>Those GRE verbal/AW scores are low even for engineering programs – probably below the 25th percentile for any top 10 school. As for GPA, the average admitted GPA for a top 10 school for the PhD is in the 3.6-3.7 range so I wouldn’t be too concerned there.</p>

4

<p>@cosmicfish: My undergrad institution is in singapore, I won’t mention the name to maintain some degree of anonimity :P. Nevertheless, my GPA probably fall in the 70-80th percentile here. Regarding the fit, all the universities I mentioned above have a strong computer vision program, so no problem there. As for the GRE, yes. I probably would retake it as my first was intended for trial anyway. Any other comments?</p>

<p>@gthopeful: Yes, I would retake the GRE. As mentioned in the starting post, I took it without preparation to get a sense of how bad I’d do. Will probably aim for 550v and 5w. I don’t want to invest too much time in it. What do you think?</p>

5

<p>gt was more or less correct on your gre - your verbal and AW would be barely acceptable at best. Most schools like to see something along the lines of a 450V, 4.5AW and if you are an international student that is even more important.</p>

<p>Your gpa does not sound great for the places you mentioned - Top-5 schools are usually pulling in students who are in the top 10% of their class at pretty darned good schools. I am not sure how well known your school is in the US, but 70-80% is not a great place to be for places like MIT and Berkeley.</p>

<p>I would still go ahead and apply to those schools - they will be “reach” schools but your research and LOR’s could still pull you through. Because of your reliance upon your research do not be surprised if they actually dig up your presentations and papers to judge your performance. I would also strongly consider a few lower-tiered schools where you stand a better chance.</p>

6

<p>@cosmicfish: Yes, I’m well aware that my overall GPA is not really that great for those kind of schools. That said, I don’t think it’s that bad either considering that my major gpa probably would still fall in the 85-90th percentile taking into account the fact that I bombed my 2nd and 3rd year in college because of my disinterest towards the subjects. What I mean is that there is clearly a trend on where I did well (CS and math stuff), and where I didn’t (hardware stuff).</p>

<p>The main reason for me in creating this thread is this question on where I’d fare with other applicants considering these facts:

  • My overall GPA is low for those schools.
  • My major GPA, although better than my overall, is probably still on the low-ish side.
  • My research credentials is above average, even for students admitted to those schools. I checked the profiles of Stanford’s grad students and probably only find 2-3 students that has about the same research credentials as me during the time they’re admitted to grad school.</p>

<p>If I already have a good GPA, I probably wouldn’t have created this thread.</p>

<p>So I guess it all boils down to this: I’ve heard that top schools are essentially looking at students as a paid research assistant. So in essence, what they look in applicants is their potential to produce innovative, publishable results. Hence, they place heavier weights on research credentials more than other things. It just so happens that many people that have good research credentials, also have good grades. But that’s what I hear anyway. Althought it seems prettly logical, it might not be necessarily 100% correct. So cosmicfish and gthopeful, if I’m not wrong you guys are grad students right? What do you think?</p>

7

<p>Sorry for bumping this thread. But I feel like threre are still a few unanswered questions that might be useful for people who are planning to apply to EECS or even engineering in general. Anyone care to responds? :)</p>

8

<p>You are mostly correct - for PhD programs they do generally value research over grades, but it is not that simple. </p>

<p>First of all, the first round of admissions is usally a downselect, where they are really just looking at the bare numbers. At some schools they will immediately pass you over - even if it is not justified - if your gpa or gre seem out of line compared to the other applicants.</p>

<p>Second of all, as much as they like your research and LOR’s, they will need to weigh in on your likelihood of completing the required coursework - if you are not solid in academics you could get kicked out before anyone cares about your research. In your case you have the major gpa going for you, but it will still be a concern - as you indicated, you did poorly in classes that did not interest you, and depending on how it appears in your SOP that could be worrying for any “breadth” courses you need to take.</p>

<p>Third of all, the schools you are talking about are super competitive. They will like your research, but will they take it over someone who did a couple of summer projects (with good LOR’s) and has a 3.8 overall gpa? I ask because those people are applying too.</p>

<p>Ultimately, you should make it to the second round at some of those schools, and then it becomes a crap shoot - some school(s) may take the chance, others may play it safe. You DO have a chance, but you need to be thinking mitigation and alternatives.</p>

9

<p>whatever happened to that article written by the CS prof that stated a 3.0 and above was good enough? i’m sure that’s still the case, especially if you come from a good undergrad program.</p>

10

<p>Did you apply at your undergrad university for a MS or PhD spot in the department you worked at as a student?</p>

11

<p>@cosmicfish: Regarding the comparison with 3.8 GPA candidates with good research experience, that’s exactly my point about how each part of the application is weighted. So it’s like their good GPA + good research experience against my below average GPA + more proven research experience. I know that there’s probably no clear answer to that but I guess I just want to see what’s the general opinion.
Regarding the schools, thank you for your suggestion. Yes, I’ll probably add 1 or 2 other schools but I’d imagine they’d also be a pretty competitive schools. I’m not really aiming for schools anyway but rather for the professors there.</p>

<p>@Batllo: No, I didn’t apply at my undergrad university for MS or PhD. They gave me an offer though, but I feel like it’s better for me to get a different exposure elsewhere.</p>

12

<p>explorer - no one can give you the weights that will be assigned to your gpa and research except the admissions committee members, and they categorically will not do so. Personally, I expect that Stanford, MIT, and Berkeley at least will turn you down, but I could be wrong - they only accept something like 2-4% of applicants, so you have to beat out 25-50 other applicants to get a spot. I think you may have a chance at UIUC and CMU, and probably at the other schools.</p>