Statistics PhD evaluation for 2022, international mathematics graduate

[SteelBite]

Hi,

I graduated with a degree in mathematics from a non-American tech school (ranked 151-200 for mathematics by ARWU) in 2018 and dropped out after a year of grad school in mathematics at the same institution. After working as a formal verification engineer for over a year, I've decided to apply for a PhD in statistics in the coming round.

These are my stats:

BSc GPA: 91.1/100 (cum laude)

Major GPA: 90.8

Unfinished MSc GPA: 97.2

GRE: 170Q, 162V, 4 AW

I plan on taking the GRE subject in mathematics.

Relevant classes:

Calc 1 (differential calculus in one variable): 88

Calc 2 (integrals and basics of multivariable calculus): 95

Calc 3 (advanced multivariable calculus): 100

Linear Algebra 1 (up to eigenvalues and CH): 92

Linear Algebra 2 (JNF, inner product spaces and bilinear forms): 92

ODE: 88

Intro to numerical analysis (mostly about evaluating functions, derivatives and integrals rather than matrix products): 96

Intro to probability (up to CLT and SLLN): 99

Real analysis: 95

I've taken many more classes in mathematics, though I'm not sure how relevant topology or abstract algebra would be.

With the exception of complex analysis, I got 83 or more in every class in mathematics.

Research:

Not much, but proved the conjecture made in my MSc for the basic cases.

Other skills:

Strong C and python skills, working knowledge of sklearn and xgboost

TA'd linear algebra for EE for five semesters, have my own PDF for it, and recognised for excellence in teaching in three of those semesters. Also TA'd several other classes.

Other remarks:

Currently going over Larry Wasserman's All of Statistics to make up for the lack of statistical background

Won and placed at several national mathematics olympiads

 

I'd love to get some input on my profile, including what I could improve. I've made a preliminary list of 15 institutions in the US, but at this point I'm clueless as to how realistic it would be.

Thanks!

Edited May 15, 2021 by SteelBite
Additional information

[DanielWarlock]

First, you need to tell us what school you went to and what percentile you ranked there in your cohort. Cornell, CMU, Duke, UPenn are all ranked at 101~150 in math but these schools are much more highly regarded compared to Hunan University, Chongqin univeristy from China which are ranked similarly on ARWU math. Whether you school is known and most importantly whether students from your program are admitted regularly to good PhD programs in US makes a huge difference. For example, indian statistical institute (ISI), Zhejiang Univeristy sent many students top phd programs but are ranked at 100-159 and ~300 respectively on the ARWU ranking. This means that if you are a top student from ISI or Zhejiang university, your application result would be drastically better than say if you are an average student from Chongqin university. I would recommend you to apply for top 50 or top 30 universities if you fall into the former category (top student from known international schools) but if you are from somewhere unknown and placed mediocre in your cohort, I would recommend you be cautious and emphasize on 50+ range.

The main problem is that you appear to have little background in statistics.  In theory, this is fine, but you will be at a disadvantage when compared to other candidates, many of whom have worked with famous statisticians and acquired in-depth knowledge of a subfield. Basically, it is difficult for you to write your SOP without actual experience working with some statistics-related projects. Say, why do you want to apply for statistics and what kind of statistics you want to do and what you know or have done about it? You could probably do some projects on statistics-related topics before you apply--even short, simple ones that take a few weeks will drastically improve your profile. Otherwise, if you like to work with probabilists, make sure to say that in your SOP and do a little bit of research of the profs you may potentially work with.

That said, you appear to have strong math background--some schools do dig that. I don't know what medals you have won but if it is something like Putnam, then it can help your profile as well. I also recommend you to take GRE math subject test. You are a math major so this should not require too much preparation for you. Getting a 90%+ rank can drastically improve your chances. 

[SteelBite]

Thank you for your input!

I went to the Technion (Israel). Most students here aren't really interested in studying in the US as far as I know, but some have gone on to doctoral programmes and postdocs at various schools in Germany, as well as ETH/EPFL. As far as North America goes, a couple of years back, one switched mid-masters to a maths PhD at Caltech, and a few PhD graduates have gone on to postdocs at IAS, Michigan and Toronto.

As the usual annual cohort of pure math students is in the single digits, mine included, I'm not sure how much meaning class rank would hold, but I did graduate cum laude.

As far as my interest in statistics goes, I've enjoyed manually compiling and deducing from data since childhood, I found the undergrad probability class easy, and the parts I like best about my job are the quantitative ones, where I have to reduce large arithmetic problems. I also have an interest in using quantitative mathods, particularly time series analysis, in order to contribute to our understanding of success in financial markets and optimising healthcare down the road. I've also learned a lot on kaggle, and would love to back up my technical knowledge there with statistical theory.

How does one go about finding a reasearch project, especially when out of school for a year and a half?

The competitions were mostly Israeli-based ones and at the high school level so I don't know if that would hold any weight, but I did get a honourable mention at the tournament of towns, which over here also serves as an IMO feeder.

I definitely plan on registering for the subject exam as soon as ETS allows it.

[SteelBite]

Update:

I've gotten my (incomplete) MSc advisor and another professor from my alma mater with whom I took a class and two seminars (alas, in algebra) to agree to write rec letters for me, and I'm pretty sure the algebra professor is going to give me a very strong letter. Would a TA with whom I took calc 1 and probability and knows me fairly well be a good choice for my third rec letter, or should I go with a professor who knows me vaguely?

I should also mentioned that I scored 83 in both optimisation and intro to applied mathematics (Fourier series, advanced ODE and dimensional analysis).

I guess I still want to understand my reach/realistic/safety range for schools given my stats.

[StatsG0d]

On 5/26/2021 at 4:51 AM, SteelBite said:

Update:

I've gotten my (incomplete) MSc advisor and another professor from my alma mater with whom I took a class and two seminars (alas, in algebra) to agree to write rec letters for me, and I'm pretty sure the algebra professor is going to give me a very strong letter. Would a TA with whom I took calc 1 and probability and knows me fairly well be a good choice for my third rec letter, or should I go with a professor who knows me vaguely?

I should also mentioned that I scored 83 in both optimisation and intro to applied mathematics (Fourier series, advanced ODE and dimensional analysis).

I guess I still want to understand my reach/realistic/safety range for schools given my stats.

It's really hard to say. Typically, a lot of non-US universities have sort of a pipeline with various US-based PhD programs. If yours doesn't have that, it's honestly a crapshoot. Your best bet is to apply to schools that tend to admit a lot of international students (e.g., University of Florida comes to mind). There are probably many others, but that's the one I know for sure tends to admit many international students. Browse this forum--I'm positive others have mentioned other programs.

[SteelBite]

5 minutes ago, StatsG0d said:

It's really hard to say. Typically, a lot of non-US universities have sort of a pipeline with various US-based PhD programs. If yours doesn't have that, it's honestly a crapshoot. Your best bet is to apply to schools that tend to admit a lot of international students (e.g., University of Florida comes to mind). There are probably many others, but that's the one I know for sure tends to admit many international students. Browse this forum--I'm positive others have mentioned other programs.

Thank you for the non-coddling response. The lower-ranked programmes on my radar are at Rutgers, NYU, USC and Boston U, and the moderately ranked ones are UT Austin, OSU, Northwestern and Pitt. At least some of these still seem to be rather selective, so perhaps going further down would be a better idea. Basically, my three main considerations are academic rigour, a decently sized metro area where one can live without a car and the size of the stipend.

[StatsG0d]

1 minute ago, SteelBite said:

Thank you for the non-coddling response. The lower-ranked programmes on my radar are at Rutgers, NYU, USC and Boston U, and the moderately ranked ones are UT Austin, OSU, Northwestern and Pitt. At least some of these still seem to be rather selective, so perhaps going further down would be a better idea. Basically, my three main considerations are academic rigour, a decently sized metro area where one can live without a car and the size of the stipend.

The problem with NYU and Northwestern is that both programs are highly selective relative their ranking. This is typically because some international students care more about university prestige rather than departmental prestige. 

The other schools don't seem bad. If being in a somewhat large metro area is important to you, you can apply to some of the larger programs in those areas such as NCSU (downtown Raleigh), OSU (Columbus, OH), Minnesota (Minneapolis, but cold). Some of the larger state schools in smaller areas tend to have really sophisticated bus systems that can get you around relatively efficiently without a car (e.g., Florida State, Iowa State, South Carolina, etc.)

You can browse through each school's / community's transit page and see where the buses go and how convenient it would be to e.g. live somewhere within walking distance to a grocery store and being able to take a bus to campus.

[SteelBite]

Updates:
Took the TOEFL, not worried about my score there. Still waiting for registration to open for the mathematics GRE.
Going to take an intro to statistical modelling class at a local university (Shanghai 76-100 subject ranking in statistics) this coming autumn semester, so sadly no grade by the time applications are due. Also considering a class in mathematical finance. Would these look good on my application? At any rate, I plan on doing a master's there if I don't get into a PhD programme.
As far as my research interests go, I'm mostly interested in applications of statistics in finance and medicine.

As far as schools and expectations go, are these realistic?
Reach:
Harvard, Penn, Columbia, Chicago, CMU, Washington, UCLA, UW, MIT (mathematics department)
Target:
UT Austin, OSU, Northwestern, Pitt
Attainable:
Rutgers, NYU, USC, Boston U

[SteelBite]

Update:
Taking classes (in particular mathematical statistics; linear and logistic models and nonparametric tests; causal inference; non-cooperative games; computational statistics), though sadly no grades will be available by the time applications close.
Does this work in my favour? Still stressing out about being an international candidate with no statistical research experience and fearing a worst-case scenario where I get rejected by every single one of the schools.