Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Courses Helpful math courses for graduate study in statistics?

  1. Sep 26, 2010 #1
    I am an undergrad planning on grad school for statistics next fall. I have the summer and next semester to take some courses. I would like to take some math that interests me but I also would like to do something that will be useful to graduate school. I have pretty much taken minimal math to get go to grad school. Here is what I've taken. Calculus 1,2,3 , Linear Algebra, A basic set theory class that is an intro to proofs, Linear Algebra, and now I am taking Advanced Calculus. Next semester I have the option to maybe take a class about machine learning or I can maybe take topology if I can get into a class, what is useful for statistics?? PDE's? More Linear Algebra?
  2. jcsd
  3. Sep 27, 2010 #2


    User Avatar

    I'd volunteer the suggestion to take a programming course. Depending on what kind of statistics you want to get into, it could come in handy.
  4. Sep 28, 2010 #3


    User Avatar
    Science Advisor

    Hello there. Just curious, have you taken any stats classes at all? Where I am (Australia) you usually are required to have a major in statistics that includes a year long A-level intro (Probability and 'Statistics') and then some subjects like Experimental Design, Markov Modeling, General Linear Models and so on. I know that you can do the 3rd year courses as graduate courses but even so the bare minimum for grad courses here are a major in stats which includes your basic math progression (Calculus, Linear Algebra, Differential Equations etc) and additional stats units. It just sounds like with the subjects you have done that you would have to do like a graduate diploma first or do some kind of transition Masters program.

    As per your question, you should definitely take a good analysis sequence and take some hard graduate courses like general linear models, markov modeling, some measure theory, and then specialist subjects depending on what you want to apply it to. With finance you will deal with the deep underlying mathematics that apply to discrete and continuous time stochastic processes, where as you would deal with Epidemiology kind of stuff with Biostatistics. Insurance has specialized knowledge as well.

    One thing I would want to say though that two-fish quant often says which I think is extremely accurate is that you will want to do everything in your 'education' phase to a point where when you end up taking employment, that you will be able to handle 'abstract' situations where you may have to do something that you weren't necessarily taught in your education phase but nonetheless have to do in your work. I'm certain that if you end up with a PhD that this issue will not exist for you. If you do however not decide to do a PhD I would probably recommend that you do a thesis of some sort that corresponds to some level of original work which will prepare you much better for work environments. I think most Masters programs do incorporate a compulsory thesis aspect, but i'm not sure if this is always the case.

    Also if you intend to do something like become a Biostatistician or an Actuary, you will need professional recognition from that industry body and thus fill those requirements. Another tip is to also learn about procedural programming and associated aspects.
  5. Sep 28, 2010 #4
    thanks for the advice. Yes, I have taken some basic stats, Linear Regressions, ANOVA and Design of Experiments, a SAS programming course, and some Mathematical Statistics. Not being out in the work force yet and having NEVER had a job in any kind of field that requires any mathematical knowledge, I am unsure of what would make me flexible in regards to different situations. The only thing I can think of that that kind of arrow would point to is to take more math classes and as much Of the rigorous theoretical side of things like probability as possible. There are 3 possible statistics master's degrees I am able to pursue, Applied Statistics, Biostatistics, and Mathematical Statistics. The applied is geared more towards jumping right into the industry and the other are more geared for those going into a PhD. Applied lacks in it's mathematical rigour from what I understand. I beleive this would make me more flexible and overall more knowledgable about the subject matter. But for my undergrad I have 1 more semester left.... Then I dont have much for choices for my classes, It's pretty much set in stone. Except for the summers...
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook