Statistics PhD Profile Eval - School Suggestions and Competitive Range

[fb1]

Type of Student: International (Canadian)

Undergrad Institution: Top 5 Canadian University

Majors: Honours program in Comp Sci. + Stats

Minor: Physics

GPA: 3.75

GRE: 160V/164Q/4.5

Research Experience: 

3 REU's, two of which resulted in honours theses (one CS thesis and one Stats thesis). Although I have no publications, one of the REUs resulted in a paper that is currently being worked on and will probably be submitted post-application, if that matters. Also, I wrote a "project paper" that uses tools from stochastic PDE's, analysis, and topology to prove the behaviour of stochastic algorithms which impressed a prof. I was thinking of creating a personal website to showcase these theses, projects, etc. 

Graduate Courses: 

I will be graduating with 8 graduate courses (2 of which are PhD courses):

Time Series(A), Generalized Linear Models(A), Mathematical Topics in Machine Learning(A), Applied Machine Learning(A), Regression and ANOVA(A), Reinforcement Learning,(*) Computational Statistics(*), Probabilistic Algorithms(*).     * denotes currently enrolled 

Relevant Undergrad Courses:  

Honours analysis I (A), Honours Analysis II (B), Honours Linear Algebra (A-) Honours Stats (B+) Abstract Algebra I (A), 

Other: 

Awarded a fellowship to conduct research over the summer.

Elected into student council twice and volunteer teaching music lessons to elementary kids.

Letter of Recommendations:

I currently have 4 professors who said they are very happy to write letters for me. Two of which said they will write "excellent" letters for me. All the profs are well-known in their respective fields. 

Notes: 

Hey guys! I'm mainly applying to PhD programs in statistics and concentrating on the following programs (they may be a reach but fingers-crossed):

1. CMU 

2. UCLA

3. University of Washington

4. Yale

5. University of Toronto 

In short, I think my weak points are my GPA and GRE. I'm hoping my letters and the fact that I performed well in advanced courses can compensate for them. I essentially completed a masters degree during my undergrad so I feel like I wouldn't get much out of applying to a masters program.

If anyone could give their opinion on schools I am competitive for and what other programs I should look into, that would be greatly appreciated! I'll take any information on biostatistic program stoo! My backup is a masters at my current institution which I am confident I can get into.  

Thanks and good luck to everyone else applying!  

Edited August 30, 2020 by fb1

[trynagetby]

Hi! I think you have a good shot at UCLA. CMU and UW are definitely reaches for most people but I think its worth applying for you. Your profile is really similar to mine so if you go look at my profile evaluation I think you'd get a lot out of it!

On a side note I think you'd benefit from retaking the GRE. Word on the street is that you want to be 90% on the GRE at least. I got a 164 Q the first time I took the GRE too, but a little studying and you can definitely get a 170 with your math background.

[fb1]

Thanks for the reply!!!

i was thinking the GRE would be weighted less this year due to the Covid situation and the unfairness of the test-at-home situation. The GRE is a pain for me to write haha.. anyways thanks for the advice and best of luck to you for your applications!!!

 

[trynagetby]

I totally agree that the at-home testing sucks and schools should just take the MIT approach and not even aceept the GRE. However, from what I know from Stats professors I've worked closely with at my home institution, they consider the GRE pretty trivial and aren't very understanding of the situation haha

Also best of luck to you too!

Edited August 30, 2020 by trynagetby

[fb1]

Also I got 166 instead of 164Q haha 

do you recommend schools like UC Berkeley, Stanford, and Columbia?

[trynagetby]

You can try but I think chances for mere mortals like us is pretty low. People I know who get into those (these are domestic student btw) have perfect grades in Analysis and additional classes like Measure theory, Differentiable Manifolds in R^n, functional analysis, and topology. But then again, if you have good grades in really hard physics classes that might help a lot. But anyhow you should only apply if you're okay with the fact that there's probably a ~5-15% chance of getting in.

[Casorati]

Schools like Stanford, Berkeley and Columbia are extremely unlikely (less than 0.1% of chance) given your profile so I wouldn't bother applying to them. CMU, Washington and Yale are reaches for you. As a Canadian citizen, you have a high chance of getting into Toronto so it does not make much sense to apply to schools below that. I would add other first-tier schools in Canada such as UBC and McGill. With that said, I wouldn't apply to UCLA unless you are really interested in some faculty's research there.

[insert_name_here]

If top 5 means Waterloo/Toronto/McGill/maybe UBC, it's worth throwing in an application to Berkeley/Stanford. FWIW, Berkeley has accepted otherwise remarkable students without real analysis, though it is rare. Columbia is less competitive than CMU/UW, they'd be worth a shot regardless your undergrad school. 

 

[fb1]

thanks for the replies!

Yeah I attend a top 3 Canadian school. Canadian PhD programs (with the exception of UofT) requires a master degree. Having taken so many master courses already I prefer direct entry to PhD hence why I am not applying to UBC/McGill. 

[Casorati]

7 hours ago, fb1 said:

thanks for the replies!

Yeah I attend a top 3 Canadian school. Canadian PhD programs (with the exception of UofT) requires a master degree. Having taken so many master courses already I prefer direct entry to PhD hence why I am not applying to UBC/McGill. 

If you go to a PhD program in the US, you may have to 'retake' some of the graduate courses that you have taken since they may cover different material so this will not save you time. If you go to UBC/McGill for a master's, you can waive the courses that you have taken and this will shorten your completion time. Even if you are admitted to PhD program at Toronto, you will still have to take graduate level probability/statistical inference and I don't think you can graduate faster compared to doing a master's at UBC/McGill first. In Canadian schools, if you get into a master's, it is almost guaranteed that you get into their PhD program so I would apply to UBC/McGill. If money is not a concern for you, I would apply to at least 10 schools.

 

Edited September 3, 2020 by Casorati