Fall 2020 Stats PhD Profile Evaluation

[SHBL]

Looking at PhD programs for Statistics for Fall 2020. Please take a look and give me feedback!

Undergrad Institutions:

                                     School 1. small university in South Korea (dropped out after first semester)

                                     School 2. another small university in South Korea (dropped out after one year)

                                     School 3. Californian community college

                                     School 4. UCSB (Transfer)

Undergrad Major: Statistics and Data Science

Cumulative GPA: 

                                     School 1:   0.0 / 4.3

                                     School 2:   2.3 / 4.3

                                     School 3:   4.0 / 4.0

                                     School 4:   3.51 / 4.0  (Major GPA: 3.68)

Type of Student: International Asian Male

Relevant Courses(Undergrad): 
Lower Div: Calculus 1, 2, 3 (A+), Linear Algebra (A+), Differential Equations (A), Intro to Proofs (A+)

                    Intro to computer science (A+), Computer Science Theory and Practice 1,2 (A+), Object Oriented Programming in C++(A), Programming for Engineers in Matlab(A+)

                    Elementary Statistics (A), Data Science Principles (A)

Upper Div: Probability and Statistics I (A+), Probability and Statistics II (B+),  Stochastic Processes I (A-), Stochastic Processes II (B+),  Hypothesis Tests (B+), Regression Analysis (A-),

                   Stats Machine Learning (B), Advanced Hypothesis Tests (A-),  Stats Data Science (B+),  Design of Experiments (A), Time Series Analysis (A-)

                   Data Structures and Algorithms (B-)

                   Linear Algebra (B+), Advanced Linear Algebra (B+), Intro to Real Analysis (A-), Graph Theory (P)

                    

GRE General Test:
Q: 161
V:  156
W: 4.0

GRE Subject Test in Mathematics:
M: N/A

 

Research Experience: 

- One year of undergraduate research student in a computer science research lab. (Worked in Deep Learning, Human Computer Interaction, and Data visualization, analysis) No publications. 

- One year of statistics research with a PhD student in Reinforcement Learning and financial modeling. No publications either.

 

Work Experience:

- One summer internship in image recognition, deep learning, and computer vision.

- One year of senior software engineer at an early stage school startup.

 

Publications:

7 co-authored math reference books for Korean high school students. (South Korea)

3 published journals to an unknown magazine in Korea.(South Korea)

 

Teaching Experience: Tutored at a city college for 2 years.

Coding Experience: Proficient in R, Python, C++, Java, Matlab. Comfortable with Linux environment, and Git.

Letters of Recommendation:  Two from professors I took class with(one math one stats).  One from co-worker/mentor (retired mathematician from U of Mich who knows me VERY WELL)

Research Interests: Reinforcement Learning, Deep learning, Mean Field Game Theory, Neural Signals Decoding and Modeling, etc

 

Programs, Schools Applying: Statistics PhD

University of Michigan - Ann Arbor
University of Minnesota - Twin Cities (Top Choice) 
Duke 

University of Washington 

University of Wisconsin - Madison

University of Utah

University of Colorado Boulder

UCI

UC Davis

Vanderbilt University

George Washington University

Johns Hopkins University

Virginia Tech

UNC Chapel Hill

UMASS Amherst

Baylor University

(Am I aiming too high?)

Please give me feedback!

Edited November 4, 2019 by SHBL

[bayessays]

I'm not sure about Utah/GWU/Baylor, but the rest of your list is unrealistic, especially schools like Michigan and Minnesota. International student competition is extremely intense. In addition to your inconsistent grades, your GRE Q score is too low.  You need to raise that to be considered seriously for a PhD in statistics as an international student.

[Stat Assistant Professor]

OP: I'm afraid that bayessays is right that your current profile is just not competitive for any of the PhD programs ranked at the level or UMass Amherst or higher (including GWU). I'm not sure about University of Utah or Baylor. You certainly need to raise your GRE Quantitative score.

If you are insistent on getting a PhD in Statistics,  your best bet is to first obtain a Masters in Statistics or Mathematics from some decently reputable university (not a Masters in Applied Statistics, Analytics, or Data Science but mathematical statistics or math with a stats concentration), perform extremely well there, and then to apply to PhD schools mostly at the level of Michigan State through Kansas State, with some schools like Ohio State, UConn, or University of Florida in the mix. Even with a Masters degree, Minnesota is probably a reach (but you can certainly try), and I wouldn't bother applying to Duke, Michigan, or Washington, as there is not really a realistic path for you to be admitted to these places.

[bayessays]

Also, your research interests are not really topics in statistics departments, so I don't think a stats PhD would be a good use of your time.

[Stat Assistant Professor]

@bayessays also makes a good point here. OP: if those are your interests, then a computer science department would probably serve your needs better. CS PhD admissions is also quite different in that they are much more forgiving about lower GPAs, and admissions is based heavily on research experience. Your research experience seems solid. If you are interested in topics like deep learning and computer vision, I would check out the sub-forum for Computer Science. If you are able to continue your current research and get your name as a middle author on a conference paper, then you might stand a fairly good shot for some decent CS programs.

You could also do a Masters in Computer Science, get some research experience there (by reaching out to a PI and asking to work for them -- some PI might be amenable to this, since you have research experience in computer vision, reinforcement learning, etc.), and then transfer to the PhD program. At my alma mater, a lot of the CS doctoral students I knew started out this way -- they started out in the (terminal) MS program, and while they were completing the coursework, they also worked/volunteered in a PI's lab, and then they transferred into the Computer Science PhD program directly. And some of these PhD students didn't even have a CS background, their undergrad major was in unrelated subjects like Civil Engineering. Just a thought.

 

Edited November 4, 2019 by Stat PhD Now Postdoc

[Stat Assistant Professor]

Check out Philip Guo's blog post about how CS admissions typically works: http://www.pgbovine.net/PhD-application-tips.htm