How to a Math program without Math degree

[zygote]

Hi,

So I'm interested in applying to an applied math PhD program somewhere .. or at least a Master's followed by a PhD.

The problem is that I haven't any degrees in the subject. I have an undergrad and a master's in computer science. My undergraduate math grades are not amazing (about ~2.75) but my CS grades are around ~3.8. My CS grades in the master's program are 4.0.

The thing is that I didn't get math when I was an undergrad, but I do now. My master's has been intensely mathematical, and I want to go on to work in mathematical statistics.

I really love working in this area. I don't have any research experience in Math, but I do have theses in computer science.

How do I get into a Math PhD program without a Math degree?

Thanks.

[nandelle]

You sound like you'd be great looking into something like data mining or statistical computing.

What statistics courses have you taken so far? Many Universities don't offer an undergraduate degree in Statistics, so many applicants have related degrees (Math, CS, Econ) with plenty of stats courses on the side.

I'd recommend (assuming you already have the pure math courses from your CS studies) to take several statistics graduate courses as a non-degree student (a mathematical statistics course at the graduate level? A calculus based regression and linear modeling class?) to demonstrate your preparadeness in the field. Then set out to rock that statement of purpose - tell them why doing mathematical statistics is what you *really* want to do.

What level of schools are you looking at applying to? Any specific schools?

[zygote]

Yeah machine learning is what I've been studying in my master's degree, that's how I discovered that I don't hate math like I thought I did 8) I've taught myself some relatively high-powered Bayesian stats (at the level of the Bayesian Choice by Robert) and machine learning stuff (variational bayes, expectation propagation, etc.) and my thesis involves applying/developing/improving algorithms in these areas.

BUT, I don't have the pure math courses sadly. I've only taken multivariable calc and linear algebra as an undergrad. I don't have e.g. real analysis, complex analysis, abstract algebra, etc. So I need to take a bunch of math courses, but I'd like to take them at grad level in the first 2 years of a PhD program rather than going through the undergrad-level curriculum first or doing a Master's.

ML & data mining are cool but I really want to learn more straight-up math so I can apply some more high-powered techniques to the stats. For e.g. nonparametrics my level of maths is simply not high enough. Variational bayes for example was a huge struggle to figure out because I hadn't done mean field theory or variational calc. in college, even though your average physics major gets that stuff quite early on.

So I don't want to fight an uphill battle, rather I'd like to be in a PhD program where those courses are intensively taught in the first two years along with grad-level analysis/algebra/etc. To me that sounds like an applied math program .. maybe a Stats PhD would be good too. The main thing is, how do I get in? Master's degree? Just go try to do research somewhere? Bribes? :twisted:

[alicealice]

Applicants coming from a machine learning/data mining background are generally viewed favourably by statistics admissions committees. You say that you wish to study mathematical statistics but that you are looking to get into math programs. Math and statistics are two markedly different fields. I would advise you to browse through the course offerings and program descriptions of some master's programs in both math and stats so as to get a better sense of which subject better fits your interests. Do keep in mind that stats programs will generally allow you to take a couple of grad level math courses as part of your degree.

Edit: mathematical statistics is statistics, not math.