Fall 2020 Statistics PhD/Master application evaluation

[EconMathStat]

Hi,

I'm going to apply to Statistics master's and PhD program this year, hopefully focusing on machine learning / deep learning. I have several concerns, including

1. changing my field from Economics to Statistics

2. did not take many courses from Statistics department

3. did not major/minor in Mathematics,

and for these reasons, I am applying to master's as well, so that I can have a better chance to get into PhD program afterwards. Any feedback would be greatly appreciated!

 

Undergrad Institution: One of the top 3 in South Korea
Major(s): Economics
GPA: 3.95/4.00, graduated with Magna Cum Laude

 

Exchange Student Program: UCLA (undergraduate, one year during junior)
Major(s): Mathematics
GPA: 3.79/4.00 (Math upper division GPA: 3.89/4.00)

 

Graduate Institution: The same university as undergrad (One of the top 3 in South Korea)
Major(s): Economics

GPA: 3.90/4.00

 

Type of Student: International Asian Male

 

GRE General Test: 

Q: 154 (65%) (trying to retake to improve the verbal score this summer)

V: 170 (96%)

W: 4.0 (59%)

Subject(Mathematics): 930(99%)

 

TOEFL: 110/120 (S 25, W 27)

Courses taken:

• Mathematics: Calculus (A+), Calculus with Vector Analysis 1 (A+), Calculus with Vector Analysis 2 (A+), Linear Algebra (A+),  Math for Econ 1 (Linear algebra/Calculus, A+), Math for Econ 2 (Optimization/ODE, A+), Number Theory (A+), Probability Theory A (A+), Probability Theory B (A), Stochastic Process (A-), Real Analysis A Honors (A), Real Analysis B Honors (A--), Topology (A-), Scientific Computing (A+), Numerical Analysis (P), Optimization (A+), Mathematical Game Theory (A+), Real Analysis A (Graduate, B), Real Analysis 1 (Graduate, A-)

 

• Statistics: Intro to Stat (A+), Statistical Methods (A+), Computer Programming (A+), Econometrics 1 (from Econ department, A), Mathematical Statistics 2 (A+), Mathematical Statistics (Graduate course from Econ dept., A+), Statistical Learning Theory (Graduate, A+)

 

• CS/EE: Software Programming (A+), Data Structure and Algorithm (A), Discrete Mathematics (A), Artificial Intelligence (A+)

 

• Courses will take this fall (all graduate courses): 1. Real Analysis B (for sure), 2. Artificial Intelligence Theory (possibly), 3. Searching and Text mining (if not take 2)

 

 

Research Experience: One research on applied Econometrics project (irrelevant to Statistics) / currently working on bridge regression

Working Experience: 3 years experience of TA (Math for Econ, Econometrics, Mathematical Statistics (graduate))

Awards/Honors/Recognitions: Honors for every semester in my home university and one Dean's list from UCLA, several merits-based scholarship

Pertinent Activities or Jobs: Statistics club (one year), data analysis assistant for the university freshmen evaluation team

 

Letters of Recommendation: One Econometrics professor (with whom I did research), one math professor from UCLA, and one Statistics professor. The first two will presumably be strong, and I am not sure for the last one.

 

 

Any Miscellaneous Points that Might Help: 

1. I got B on graduate Real Analysis A (measure theory) from UCLA when I was junior. I believe I was not mathematically-matured enough at that time, and I did kind of retake the course in my home university and got A-. And I am going to take Real Analysis 2 this fall to complement B on measure theory, even though I am not sure if it can fully compensate B on the measure theory.

2. I got P on Numerical Analysis just because it was my undergraduate complement course and I was not allowed to take it with letter grade. The class rank was 1/37.

3. Even though I am currently in Economics department as an graduate student, I have not been taking Econ courses for a while. And I am not getting MA in Econ here. I just happened to have a chance to take three semesters without tuition in Econ graduate program in my home univ., and I was complementing math and CS/EE courses during the this period.

4. I actually applied to 8~9 Statistics Master/PhD programs last year and got all rejected. Most of them were top schools like Stanford MS or CMU PhD, but it also included non-top schools in Statistics like UCLA MS. I had pretty similar profile except that my GRE subject math was 830 (83%) at that time, which I submitted to all of the schools I applied.

 

 

Program: PhD/Master in Statistics (and possibly OR MS)

 

School List:

 

• Master's in Statistics: UC Berkeley, U of Chicago, U of Washington, U of Michigan, UCLA, UCSD
• PhD in Statistics: NCSU, U of Wisconsin, ISU, PSU, Purdue, UCD, Florida State U, UC irvine, UCSB

 

I don't know which schools are my reach/match/safety. I am also thinking to apply to Georgia Tech OR Master's and USC CS Master's if it fits my profile better, since OR and CS also work on machine learning / deep learning.

 

Thank you for your time and advice!!

 

Edited July 24, 2019 by EconMathStat

[Stat Assistant Professor]

You may want to have one of your LOR writers point out that you retook measure theory in your home institution and got an A and explain the P, as their words carry more weight. This should also be mentioned in your SOP (though I wouldn't say it was because you were "mathematically immature," but put it in a more positive light, like how you were determined to prove to yourself that you could do well in measure theory and were successful to that end). Also, make sure that your letters of recommendation are *very* strong. These make a big difference in getting your application into the "Accept" pile.

Your profile looks pretty good, in my opinion. Usually international students from the top 3 schools in South Korea do pretty well in PhD admissions, but the top schools will of course have very, very fierce competition so it helps to apply more broadly. Your list of schools looks very reasonable. You would probably also have a good shot at schools like Texas A&M and other schools of that caliber.

Edited July 24, 2019 by Stat PhD Now Postdoc

[EconMathStat]

@Stat PhD Now Postdoc Thank you so much for your valuable advice! But can I ask if there is any reason to pick Texas A&M in particular? Or did you just mention it based on its ranking?

[GTA357]

I'm honestly surprised that you didnt even get into non top MS programs in statistics last year. On paper your profile looks good but there might be certain unobservables like references that could be hurting you.

[Stat Assistant Professor]

4 hours ago, EconMathStat said:

@Stat PhD Now Postdoc Thank you so much for your valuable advice! But can I ask if there is any reason to pick Texas A&M in particular? Or did you just mention it based on its ranking?

I based my assessment on the rankings, the fact that it is a slightly larger program, and the fact that some of the alumni I'm familiar with from TAMU are also from the top universities in South Korea (Yonsei, etc.). I don't think the fact that your degree is not in Stat is really a problem, as long as your mathematical background is strong -- there are definitely PhD students in Statistics (including international students) who come from an Econ background. But it may help to have your letter writers stress that you took a ton of advanced mathematics like probability theory, measure theory, topology, etc. You can ask them to mention this.

Edited July 24, 2019 by Stat PhD Now Postdoc

[EconMathStat]

@GTA357 Yeah, I think B on measure theory might be the reason... Hope 930 on GRE subject math somehow cancel this flaw out.

 

@Stat PhD Now Postdoc Thank you so much for your help!! I will ask letter writers to mention that!