5

I’m an incoming master's degree student at a top 10 computer science (CS) program. I’m applying for PhD programs. I think that people doing ML/AI have an unfair advantage when it comes to the number of papers they can publish.

My research has been mostly in the area of theory and algorithms, where you either come up with a proof for your theorem/algorithm or you don’t. I think this is significantly harder than what most undergrads do in ML/AI. I have seen tens of students with publications in ML, but 0 students with publications in complexity theory, for instance. I know that at there are people who have published papers in theoretical CS, to but my point is it’s rare (and hard).

Do PhD admission committees at the top 20 CS schools take this into account when they look at a student’s previous research experience? More generally, how is a student’s research experience evaluated?

Improve this question
1
  • 3
    It depends on the person: some take this into account, others do not. In any case, it should not effect how you approach the application process.
    – Moishe Kohan
    Sep 16 at 23:35

2 Answers 2

Reset to default
19

Ah, so you've noticed that people in AI/ML publish a lot of junk papers in bulk? Yeah, university professors in CS have noticed that too.

But also, some of the people who publish those sorts of papers in bulk may also be evaluating your application, and maybe they think the coolest best researchers are the ones that just publish the most papers. And also some university systems especially in certain countries favor "quantitative" measures of research output that also value quantity over quality.

So, really there is no one answer to your question: anyone who evaluates your application will do so according to their own personal criteria of what is valuable. You'll apply with the research experience you have and in an ideal world you'll be admitted to a program where other people there also value the type of research experience you have, and it'll be a good fit for both.

Improve this answer
3
  • “ But also, some of the people who publish those sorts of papers in bulk may also be evaluating your application” even at top 20 schools?
    – StackExchanger
    Sep 16 at 23:54
  • 20
    @StackExchanger Maybe even especially at those schools, I don't know. I think it's a mistake to assume "top" schools are somehow immune from bad academic trends.
    – Bryan Krause ♦
    Sep 17 at 0:52
  • 4
    @StackExchanger top schools have even a higher incentive for producing many publications, as the number of publications by faculty is one of the numbers that defines what a "top" school is.
    – Ander Biguri
    Sep 17 at 11:46
6

In both Mathematics and CS, it is well known that people in certain areas publish more than in others. (I assume this is true.) There is no opprobrium attached to working (or not working) in one of these fields, but publication records are weighted by the subfield. Taking on a Ph.D. student is such a commitment that you can expect people to actually look at publications and not just count them. Not being in a "hot" field where people only publish after lots of work might be to your advantage, since potential Ph.D. advisors in that field are not swamped.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .