5

I ran across this thread, in which the accepted answer mentions to "[..] Start on a research project (independently or otherwise) if you have not already ASAP. "

I'm an undergraduate studying mathematics and physics. I have an Associate's degree in Computer Science, and am very comfortable with programming. I'm currently taking a course in numerical analysis, and this has really motivated me to apply things we're learning to problems I come across while programming.

As an undergrad, with a lack of proper training/education to perform real mathematical research, would it look good (to future graduate schools to which I will apply) if I, for example, write a paper or two based on things I do applying mathematics to Computer Science and programming, e.g.- solving optimization issues in GPU processing of large data sets, etc?

Even if the papers aren't so interesting, would it at the very least look good on my "resume", so to speak?

For the record, I'm hoping to go to graduate school to study mathematics, not computer science, which is why I can't quite answer this question myself.

7

Undergraduate research is whatever research that you do as an undergrad, period. So if you do something that ends up being publishable, or feeds into another research project in a real way, it's undergraduate research.

However, for practical purposes, you want to get some faculty involved in your research project. It will help with funding for travel (you do want to submit to conferences rather than journals in CS), it will make the work better, and you will end up with a good letter writer afterwards. And if you think your project won't be interesting to faculty in your current school, what makes you think it will later on swoon an admission committee?

would it look good (to future graduate schools to which I will apply) if I, for example, write a paper or two based on things I do applying mathematics to Computer Science and programming, e.g.- solving optimization issues in GPU processing of large data sets, etc?

Sure - if the paper is good. Solving optimization issues does not sound bad at all to me on first glance, but of course it is impossible to tell without seeing the paper or knowing in much more detail what you would want to do.

Even if the papers aren't so interesting, would it at the very least look good on my "resume", so to speak?

I guess that depends on what you mean with "so interesting". It does not have to be A* / Transactions level quality, but the paper certainly needs to be published in a reasonable, peer-reviewed, non-spam venue. Papers in pay-to-publish OA journals don't help at all.

This seems to me to be an unfortunate side effect of the "prior research experience required" mantra that many admission committees nowadays seem to have - many students seem think that sending crap papers to pseudo-venues will help them get accepted. They don't. A paper that is substantially below the level that a prof. would want to submit himself won't help you.

For the record, I'm hoping to go to graduate school to study mathematics, not computer science, which is why I can't quite answer this question myself.

If you want to continue in maths, a maths paper will help more than a CS paper. Still, a good CS paper will be much better than no paper at all, I suppose.

  • thanks for the answer. BTW, what I meant by "not so interesting" - I'm mainly wondering that if simply the concept that I have written a paper or two as an undergrad will look good to an admissions board, despite whether or not the content of the paper makes their jaws drop (undergrads simply usually don't have that kind of knowledge yet!). So basically, a paper that a professor would say "yes of course this is obvious", but the fact that a student is trying to write papers would impress them ? – galois Feb 19 '15 at 7:11
  • 1
    @jaska The fact alone that you have published "something" will not impress them. Nowadays, most PhD student applications I get from the middle and far east have 3 - 10 "publications" in spam journals to their name. This weakens their applications, as it makes me afraid that their idea of doing research is sending material that is too weak for a class paper to a pay-to-publish journal. TLDR: quality matters. – xLeitix Feb 19 '15 at 9:45
  • 2
    @jaska Also, your paper does not need to be something that they themselves wouldn't have been able to do, but it should be something that not every other applicant could have done. – xLeitix Feb 19 '15 at 9:47
5

Let me question your question. First of all, note the disclaimer in the answer you refer to that that user just gave some second-hand advice. Also, the answer with the most votes places much less emphasis on undergrad research.

For admissions to most grad schools in math, undergrad research is neither necessary or expected (particularly published research--in fact, publications in good journals in math usually take a long time to get accepted, so the chances are low that you would have a good paper accepted by the time you apply for grad schools).

If there are specific problems that you want to research on your own, that's great, and as xLeitix says you should try to get a faculty member to help guide you. If however you just want to write a paper to help get into grad school, this may not be worth it. While research experience is good for your application (though may not be extremely helpful if it's not in math), you need to consider the opportunity cost.

What (let's say US PhD) admissions committees are looking for is evidence that you're talented, motivated and have solid preparation. Even if you've done some research, unless it's outstanding, if you don't have much advanced coursework, or didn't do well in abstract classes, you're probably in a worse position than someone in the opposite situation. The best preparation for grad school is taking a lot of advanced math classes and doing really well in them.

That said, it's great if you do your own research on the side if it's not at the expense of your coursework, and if you have a faculty to discuss things with, they can write you a recommendation letter that can mean a lot, whether you've published, or even submitted, something or not. However, if you can do something like an REU (in math), that will be even better.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.