Profile evaluation and advice for PhD in Applied Math (&/or) Statistics

[Enigmatic]

Type of Student: International (India), Male

Undergrad Institution: Indian Institute of Technology
Major(s): Electrical Engineering
GPA: 9.56/10

Research interests: Optimization, compressed sensing, random matrix theory, stochastic programming

Though I have a degree in EE, my true love has always been in mathematics. My deepest interests at the moment are at the intersection of optimization and statistics, bent towards the theoretical side. I have identified potential advisers who are distributed across statistics, math, and some in EE. My real question is on which program to apply, given that I have the freedom to work with advisers in other departments.

-- EE - I totally hated my undergrad education, other than signal processing. I have made up my mind to have nothing to do with the area, since the questions that interest me really don't have any connection to EE, other than an indirect link through optimization and random matrix theory. So I figured I might as well get the hell out.
-- Statistics - I like the programs, but the catch is, they are usually outside the school of engineering. So I am not sure about my admission chances. Also, I don't think I have a wide enough exposure to statistics, and I might have to complete a hell lot of prerequisites, making for a miserable first year.
-- Applied Math - I had initially planned for this and was filling out the applications. But some of my letter writers warned me about a few things which I found troubling. I heard that a lot of applied math programs view math and computation strictly as a means to study natural sciences (physical, chemical, or biological phenomena). Hence the emphasis is primarily on differential equations and dynamical systems, from which there is no escaping. Given my interests, would I be a misfit?

My advisers asked me to consider Operations Research programs. What is the general consensus about them? I heard from a few friends and students currently in a few such programs that OR is looked down upon by most. Essentially, depending on the problem domain, OR plays the kid brother role to either CS, EE, Stats, or Math; and given an opportunity most would choose one from the latter set as opposed to OR. How true are these comments?

Thank you so much for the help

[airy]

I think these comments are unfounded.. For example, I think the research being done at MIT's OR department (http://www.mit.edu/~orc/) is really superb and if one is interested in optimization, I think it would be an excellent choice. I really do not see why doing one's degree in an OR department would be looked down upon. And especially given your background and interests, I feel that OR might be a really good fit.

[Enigmatic]

Thanks a lot for the reply. I do feel that very specific OR programs (namely MIT, Princeton, and maybe Cornell) are good fits and absolutely awesome. It's just that I wanted to ask in a global forum (I have access to only a biased Indian picture) about opinion on PhDs in operations research.

MIT and Princeton are certainly in a class of their own as far as optimization and stochastic processess is concerned, but quite a few told me that most other programs aren't competitive. Added to my worries is the fact that nearly half the faculty in OR programs have a PhD in one of math, CS, or EE; which probably indicates some hierarchy when hiring is concerned.

Any thoughts on applied math is also much appreciated. Do they stress of applied math inclined towards optimization, stochastics, or signal processing; or are they all mostly to do with math and simulation for natural sciences? Thanks

[Enigmatic]

Bumping up the thread. With deadlines fast approaching, any comments or suggestions will be greatly appreciated. Thanks in advance.