Seeking advice on applying to PhD programs in Statistics (with a background in CS)

[approximatelyNormal]

My info:

BS: Computer Science (finishing in spring 2013)

Minor: Statistics

GPA: 3.75

GRE: 162 verbal/165 quant

A little background...I fell into computer science in my third year of college. Before that I had no clue what I wanted to do because nothing interested me. I enjoy the logic and problem solving aspects of CS more than anything else. But when I got into some stat courses I found that I'm much more interested in studying statistics, because it actually motivates me to go do research rather than simply finding the coursework interesting.

I want to pursue a PhD in statistics, but I'm not really sure what area of statistics I'd like to study. Machine learning and data mining both sound fascinating to me, but I don't know a lot about them. What are some good programs for those areas, or any other areas where a CS background would be especially useful?

[sisyphus1]

hey - also interested in machine learning/data mining (they mean the same thing) here.

some good programs would be:

stanford

berkeley

CMU

duke

michigan

most stat departments will have at least one or two professors working on machine learning these days, given that it's such a 'hot' area.

having said that, try to get some machine learning related research under your belt before you apply to these programs.

[cyberwulf]

A CS background is useful in most areas of statistics these days, since the vast majority of modern research incorporates computer simulation. Areas which make extensive use of computation include: statistical genetics / computational biology, graphical models, functional modeling, imaging, and spatial statistics.

You might also consider biostatistics departments, many of which also do a lot of machine learning and data mining work. I assume you haven't done a ton of math coursework beyond the usual pre-requisites, and this will be less of a limitation in biostat than stat.

[laplace]

well for CS + Stat, obviously machine learning and its applications are the first to come to mind like computer vision, computational biology, natural language processing and so on. However, the choice of a program will depend on whether you favor CS or stat more, so a program belonging to CS will probably focus more on the "computational aspects" of machine learning rather than statistical foundations and will be more heuristic with an engineering flavor. In Stat departments, research is more rigorous with more math flavor.

in addition to machine learning, there is an interesting area called computational mathematics, not related to statistics but might interest you.

good programs in CS with research in machine learning include: CMU, purdue, Berkeley

Edited October 16, 2012 by laplace