Profile evaluation and advice. Mathematician to statistician.

[silverss]

So I am currently halfway through my first year in a 2 years masters in math with probability theory as my specialization. So in my academic career thus far I have focused quite a bit on the theory side, but I have recently decided I would like to move to a more applied field such as statistics. I graduate from my current masters program in Summer 2016 and am trying to figure out what my next steps should be in order to plan for the change in fields. My end goal is to do research in an industry setting, preferably outside of north america. Data science/machine learning are somethings that seem particularly interesting to me right now. 

 

I have the following questions that I would love to get some feedback on if possible:

1) With my goals and history, should I go for a masters or PhD,?

2) How difficult would it be to make the transition/what should I be doing. I am worried that because of my poor background in stats and programming will set me back.

3) If decide to go for  a PhD I would like to take gap year to travel since I don't want to do 10 years of school straight. Will doing that hinder grad school prospects?

4) Realistically what kinds of schools do I have a shot at/should look into? Would I be competitive for tier 1 schools?

 

I have tried to be as specific as possible with my background.

 

Undergrad

- University of Waterloo

- Double major in Pure and Applied mathematics

- cumulative GPA 3.9+

 

Masters

- University of British Columbia

- Probability theory

- 4.0 gpa

 

Courses/Math background

- I have taken over 40 math/physics courses in my undergrad, with many at the graduate level. Beyond the core courses I overall have a very strong background in math, physics, especially when it come to analysis, differential equations, differential geometry, probability/stochastic processes.

- During my masters I have taken courses in advance courses in pdes, multiple in probability, ergodic theory. I plan on taking some stat courses and more probability, and probably a computation methods course.

- Although my probability background is very strong, my current stat/computational background is poor to say the least. I have taken the basic stat course at Waterloo as well an applied linear models course.

- I have had exposure to R, mathlab, maple, and am currently learning python, but my background is poor to say the least.

- I plan on taking the math gre next month.

 

Research/Work Experience

- I have done 2 research terms in pure math in fractal and differential geometry respectively.

- Doing a masters thesis, probably on stochastic pdes

- Did one 4 month internship as actuarial analyst (have passed first two exams).

- Currently an instructor for a section of a calculus 2 course

- Been private tutoring for 4 years at university level

 

Thank you in advance, any advice would be appreciated.

 

[epimeleia_heautou]

I only just finished the application process, so my perspective is limited. However, I too was switching fields and came with much more of a pure maths background, and I'm happy with how the season went. Here are my observations, for what they're worth:

 

1) A lack of background in statistics is apparently by no means a deal breaker. It seems that many adcoms are just as concerned if not more so with whether or not a given applicant can handle the math required for statistical theory than they with the extent of the applicants background in statistical coursework or research. The extent of my formal "statistics" education is *very* limited, and I have little to no background in programming/computation, but I have a strong maths background from my undergrad major. Adcoms didn't seem to hold this against me.

 

2) By the time I join my PhD program this coming fall it will have been over three years since I received my BA. I don't have a masters, and most of the time I spent between completing undergrad and applying for grad school had been spent doing things very unrelated to statistics. Adcoms didn't seem to hold it against me. That being said, two caveats: (1) I managed to do some things related to stats as I was applying last fall, and this helped give my letter writers recent data points. If you're intend to take time off, I'd consider asking the professors who you intend to write your letters of recommendation to draft something this year while you and your accomplishments are still fresh in their memory; (2) the volume of the applicant pool to statistics graduate programs is growing quickly in size and in depth, and it's not clear what "competitive" will mean in two or three years time.

 

3) As far as MS vs. PhD goes, my understanding is that this depends on what sort of work you want to do afterwards. It seems that if you want to be designing studies, leading projects and developing new methods for your company/team, then you'll probably want a PhD. But, if you're happy to work in less of a leadership position, an MS will suffice. You should try talking to some people who are currently doing what you would like to do. You also may find that an MS in statistics on top of your MS in maths is unnecessary -- I'd imagine you'd be able to find work as entry-level or higher analyst if you continue to teach yourself basic stats and computational programming/scripting, find some sort of more or less "applied" project to work on during the second year of your masters, and perhaps do an internship this summer.

[Stats2015]

I agree that most schools are probably more concerned with your ability to handle the math theory in the courses than that stats classes so I think you're sitting pretty well in my opinion.

I also don't think taking time off will hurt you but I would try to focus on doing something that you can add to your CV during your gap year.

I would suggest trying for a PhD mainly just for funding as most masters aren't funded. But ultimately it comes down to personal choice. The way it was explained to me is that the only difference from a MS and a PhD in industry is freedom in what you're researching. I was told that most masters sometimes end up doing "grunt" work while a lot of times Phds are the ones who are more "in-charge" and can be more involved with influencing what projects are worked on. But this is all just what I've been told.