1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!


  1. Jun 2, 2013 #1
    I'm applying to two different universities, each of which has a top-notch department doing the type of research I'm interested in, neural network modeling. One is a DARPA funded program in an applied math department, and the other is a program in an electrical and computer engineering department. The good news is that I personally know the directors of both departments, and they are both familiar with work I have already done in the field. Each of them have indicated that they would like me to apply to their programs, but I have stalled them over the past year while I've tried to get my math skills up to a less-than-humiliating caliber for the department titles. Each of these guys knows that my math skills are not what I would be bringing to the table in a cooperation, but I have to take the GRE for general admissions requirements. I just took my first practice test and pretty much bombed the quant.

    The funny thing is that my calculus skills are actually pretty good, and our modeling is mostly with ODE's and most of that is already incorporated within a matlab NN toolbox that I already know how to use. That's why they're not so worried about my math skills. My problem is with the "Bob and Sally can paint each paint a fence in one and 1/2 hr, respectively," kind of word problems my brain just does not want to play ball with. I mean, I'm trying, but I just took a practice test and didn't do so hot. It was just one test, and I was tired :redface:, but I don't even want to take another practice test much less the real one. However, I'm worried about sending these universities a low GRE score.

    So my question is, do I stall my future a little longer in order to get better at the Bob and Sally questions, or do I just go and take the damned thing and quit stalling? I mean, I guess I could take it and not send it if I don't like the score, couldn't I? Maybe not. I'm not sure how to proceed. Any ideas?
  2. jcsd
  3. Jun 2, 2013 #2


    User Avatar
    Homework Helper

    It sounds like you may have gaps in your math knowledge/skill. If you can give a list of what they are, we can help.

    PS. I mean, what is it about the Bob and Sally example that is difficult? And what other problems are difficult?
  4. Jun 2, 2013 #3
    It's not so much that they're difficult is that I resent feeling like I'm being treated like a child by having to solve these little stupid riddles. I just feel its kind of a slap in the face after all the hours of my time I've given to refereeing articles for various journals, one of which I sit on the editorial board for. So now, for some reason at my age my worthiness as a member of this community is going to be judged on how I do on solving some remedial puzzles?

    The way I look at it, the reason standardized tests such as the GRE are given in the first place is to try, in some way, to get some kind of a hint as to whether that individual has the intellectual acumen and perseverance skills necessary to do research and publish in a given field. What other reason could it exist for? So now you take someone who is established in a certain field and then make them go back to school to make sure they know their ABC's? Well, my ABC's are a little rusty compared to some 22 kid in his senior year of undergrad. My XYZ's, though are just fine. I haven't used my ABC's in 20 years because I'm doing more advanced things now. But now I have to take several months out of my life to re-learn them for prime time? And why? So you can get some kind of a hint as to whether I have the acumen and perseverance to do research and publish in a field? See the circular nonsense here?

    In any case, it is what it is. I've done a measure of study for it and answered about 2/3rd of them right. I can study for another month and get maybe 3/4ths right, another month and get 4/5ths, etc. I don't need help with the problems I just need to take it and get it out of my life. My question is will it hurt me to expedite that, or should I wait the month or two for the better score?
  5. Jun 2, 2013 #4


    User Avatar
    Homework Helper

    If you're already comfortable with the "XYZ's", why will it take you several months to master the "ABC's" again? Even if there are some things you didn't learn 20 years ago, are you not confident that your experience would enable you to pick up the 'basics' quickly?

    It sounds your pride may be impeding you from doing well on the GRE. If the whole time you're studying you're thinking that it's a complete waste of time, then yeah, I can imagine that you are not as focused as you should be and are thus prone to making mistakes, resulting in you doing poorly on it.

    Have you asked the directors of the prospective departments how your work and experience so far would weigh against your general GRE score? Their opinions would weigh more than ours here. It's possible the departments will let you in with a lousy GRE score if they know what you're capable of. It's possible that a low GRE score could signal to the rest of the committee that you're not willing or ready to jump through the tedious hoops you'll have to go through in grad school.

    If it's the latter case, you may need to just swallow your pride, stop thinking of the problems as stupid riddles, and study. It sucks, and it's a blow to the ego, but if it's in the way of you getting what you want, then you have to do it.

    Hopefully your previous experience will outweigh the GRE, though you should still treat it seriously and study. I wouldn't delay it by several months - you have to take it at some point - but another month, maybe two, of serious studying should help you bring up your score, hopefully more than the incremental amounts you seem to think it will.
  6. Jun 2, 2013 #5


    User Avatar
    Homework Helper

    In advanced driving courses, one of the things that is taught is to separate tasks. For example, when turning a corner, one should choose the correct gear before starting the turn. And, one should not be using the brake when entering the turn, the braking should be done already. Turning while braking is bad practice, as is changing gear while turning, unless it is a long gentle curve.

    Why do they do this, and why is it considered advanced? Because, if each task is separate, each task is always the same to execute and nothing can get muddled up. They also teach breaking down complex tasks into simpler ones. Overtaking is a complex task and is broken down into simple phases that don't overlap, making the whole endeavour entirely predictable and repeatable.

    Likewise in math, if one has a repertoire of simple techniques mastered, complex problems can be broken down into their elements and novel situations are not as difficult as they might be with an ad hoc approach. So I would say that one of the reasons for a test like the GRE is to see who is good at mastering many simple techniques, and those tricky questions are there because one is supposed to break them down into those elements that one has mastered. Tricky questions thwart ad hoc attempts at solving them, to catch those out who have not mastered the elements.

    And I agree with them that it is important.
  7. Jun 2, 2013 #6
    Snobbery and a sense of entitlement, make sure to really embellish those qualities in your application because that's exactly what an employer is looking for
  8. Jun 2, 2013 #7
    Thanks Mute, you're 100% right. It really is a projection of my failure to have taken the test back when I should have, along with applying to grad school back when I should have. And I'm taking it out on the poor test because its too painful to look into the mirror. You're also right in that it is my pride that is the major challenge I need to overcome, a far more formidable adversary than the test itself. Maybe I just needed to hear that from someone else, so thanks again.

    @ Verty

    Nice analogy...point taken, thanks.

    @ Gauss M.D.

    I assume your being facetious or else I'd be a shoe-in for the job:approve: But again, point taken.
  9. Jun 2, 2013 #8
    1.) The ability to solve "word problems" is an essential skill for anyone in a technical field because it is the test of your ability to apply mathematics to solve real worlds problems.

    2.) I find the "our modeling is mostly with ODE's and most of that is already incorporated within a matlab NN toolbox that I already know how to use" statement troubling for someone perusing education. It seems like that is the hobbyist attitude. You don't need to understand the details, just let tools do the dirty work. I would expect a researcher to try to understand each stage in detail.
  10. Jun 2, 2013 #9
    Well, I know the maths and the implementation of it pretty well. It is basically a system of coupled oscillators modeled by coupled sets (lots of them) of 2nd order non-linear ODE's, with values assigned to synaptic weights to define the system parameters. I developed derivatives of the main framework that were more specific to a model I was developing in the late 90's. The qualification there is that, at the time, I learned the math's to use and develop the toolbox specifically for that project, so in that sense, my knowledge and experience is myopic to that project. That is why I'm going back these days to try and shore up my foundational maths. Right now it is through self-study, but I'm hoping to get a belly full in grad school.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook