2019 Statistics PhD Profile

[galois]

Undergrad Institution: Top 30 Liberal Arts College
Major: Mathematics

Concentration: Scientific Computing
GPA: 3.91

Class Rank: Top 4%

Type of Student: Domestic Mixed-Race Male

Relevant Courses:  Calculus I-III (A), Intro Statistics (A), Random Structures (B+), Foundations (A), Linear Algebra I-II (A), Abstract Algebra I-II (A), Number Theory Seminar (A), Mathematical Logic Seminar (A), Complex Functions (A), Real Analysis I (A), Euclidean & Non-Euclidean Geometry (A), Intro to Programming (A), Data Structures & Program Design (A), Complex Systems in Scientific Computing (A), Scientific Computing Seminar in Parallel Computing (A), Independent Study in Computational Algebraic Coding Theory (A)

 

GRE (V - Q - W):  167 - 168 - 4.0  (98% - 94% - 59%)

GRE Subject Math: (taking 9/15 for the first time)

(Note: I'm cramming now, but given the time since undergrad, this subject score will likely be horrendous. Only taking for UCSD / Yale.)

 

Research Experience: 

- REU in combinatorial mathematics: presented at JMM, won poster presentation award, resulted in publication

- REU in gravitational physics: worked in a lab in Paris doing some experiments and analysis on a table top simulator for LISA mission.

Teaching Experience:

- Lead tutor for Calculus I-III for a few years, also helped in class with Maple syntax and stuff.

- Tutored an autistic client over a summer and increased his placement exam scores so that he could attend the local community college program.

Coding Experience: I've been in software development 4 years. Standard web technologies, passable Linux sys-admin skills, experience in PHP, Python, Javascript, Bash, etc., and more recently functional programming in Haskell. Did some C, C++, Maple, Mathematica, Matlab back in college as well.

Recommendation Letters: Although its been 5 years, I made a great impression on my profs, and they were very happy to hear that I've finally decided to go back to school. I think these letters will be very strong. All 3 from Math department.

 

Programs Applying:  Mostly Statistics PhD.

Research Interests: I'd like to be able to stay within the realm of the software industry, but move away from computer science engineering problems and get closer to mathematical problem solving, ideally in a research role. I think the big data trends have opened up these sorts of opportunities in the form of research scientists dealing with ML / computational statistics / etc. in industry. That's kinda my main motivation for pursuing graduate school, so I'd like to have that kind of computational focus, high dimensional data, etc.

 

Schools of Interest (Statistics Degree unless otherwise noted):

UC Berkeley

Caltech (Computing & Mathematical Sciences)

Carnegie Mellon (Statistics with ML Joint degree)

University of Washington

Duke

UNC Chapel Hill (Statistics & Operations Research)

UC San Diego (Mathematics with a Specialization in Statistics)

Yale

UC Davis

USC Marshall School of Business

UT Austin

UC Irvine

UC Santa Barbara (Statistics & Applied Probability)

 

I think I've finalized this list of schools. But honestly I have no idea if I'm off base, even after looking at everyone else's profiles. Am I throwing money away by applying to the top tier like UCB, UW, CMU, etc.? If you have suggestions for adding/removing programs, please comment! Would love to save money on applications. My main criteria: 1) temperate climate, 2) affordable living, 3) strong theoretical coursework 4) alignment with research interests described above, and 5) small-ish departments.

 

In addition to recommendations, does anyone have experience with the programs at UCSD or UC Irvine? UCSD is notably missing from Stats ranking but its Math and CS programs are very well regarded. Does anyone know what the stats specialization is like? Also, the Irvine department website seems half developed, so it's hard to get a feel for that department.

Edited August 27, 2018 by galois

[bayessays]

You're definitely not wasting your money applying to top schools. If anything, I would reconsider some of your options and what you are looking for in the type of department (you can do better than UCSB and USC, UCSD has one really good stats person I can think of but isn't a stats department and will have very different focus) - you're missing schools in the 10-30 range where I think you would have seen good success, but you're limiting yourself geographically so that takes out some options. Have you looked at UCLA and NCSU? I think those are good target schools in warm climates. 

[galois]

Thanks for the advice. It has been tricky finding schools with such strict geographic requirements, most of which are imposed by my fiance - she really hates the cold. But, since we're moving solely for my education, it's only fair. It's a shame so many of the good programs have such insane costs of living though. I'm afraid of actually getting into Berkeley and then having to figure out how to afford it?.

I did have UCLA and NCSU on my list at one point. I read some fairly damning things about the UCLA department on this forum, but I suppose those were just anecdotes. The reason I took NCSU off the list was mainly its size. I like a smaller student-faculty ratio. But if I'm shuffling things around, removing some of those UC schools, I should reconsider it.

[bayessays]

Also, to address your other question, UC Irvine is a newer department but they have some really great professors. Some other schools in warmer places I can think of for you: Rice, Florida, Florida State, UNC Biostat (not sure if they'll let you apply to both, but their biostat department has good ML people and is very mathematical).

[galois]

OK, thanks for your help @bayessays. I think my list is final. It has actually changed quite a bit in the last week or two, influenced by your comments and further research.

1. U Washington - Statistics
2. Caltech - Computing & Mathematical Sciences
3. Duke - Statistics
4. UNC Chapel Hill - Statistics & Operations Research
5. NC State - Statistics
6. UCLA - Statistics
7. UT Austin - Statistics
8. U Florida - Statistics
9. UC Irvine - Statistics
10. College of Charleston - Mathematical Sciences with Statistics Concentration, MS

I feel good about this for the first time since starting the search a few months ago. Thanks again.

[bayessays]

You ever think of biostat programs? particularly for Washington, which is probably one of the hardest stats programs to get into - their biostat program might be a little more attainable and has amazing ML people (eg Daniela Witten).  I think you have a good shot at getting into some program on your PhD list but they're all pretty competitive programs. Good luck.

Edited September 7, 2018 by bayessays

[StatsG0d]

2 hours ago, bayessays said:

You ever think of biostat programs? particularly for Washington, which is probably one of the hardest stats programs to get into - their biostat program might be a little more attainable and has amazing ML people (eg Daniela Witten).  I think you have a good shot at getting into some program on your PhD list but they're all pretty competitive programs. Good luck.

Yeah, I would apply to at least a couple biostat programs in the top 10 pooled. If you're concerned about rigor, Washington and to a lesser extent UNC have a strong theoretical focus. 

[galois]

50 minutes ago, footballman2399 said:

Yeah, I would apply to at least a couple biostat programs in the top 10 pooled. If you're concerned about rigor, Washington and to a lesser extent UNC have a strong theoretical focus. 

When you refer to UNC having a theoretical focus, is that strictly for the Statistics degree? I was also considering the Statistics & OR combined degree, but after comparing the curriculum difference, I'm concerned the latter will have less theory and be more applied.

EDIT: Just realized you were referring to the Biostat degree. Thanks for tip. Do you have an opinion on the Stat/OR degrees though?

3 hours ago, bayessays said:

You ever think of biostat programs?

I have mixed feelings about this. Everyone on this forum, and elsewhere on the internet, is of the opinion that they are pretty much the same degrees. But, while my interests in statistics itself is rather broad (high dimensional data analysis/inference, ML, topological perspectives, probably more that I'm not aware of yet), I am picky in my applications. For whatever reason, I'm not particularly interested or inspired by biology as a science. I am aiming to apply research in the area of environmental science, ideally climate/weather forecasting, possibly another physical science like astronomy. I would also be interested in sociology applications, specifically regarding economic inequality, effects of public policy on such problems, etc. So, to that end, most of the programs I have picked have at least some faculty applying research to those domains.

What do you think? Are there flaws in that thought process?

Edited September 8, 2018 by galois

[StatsG0d]

24 minutes ago, galois said:

When you refer to UNC having a theoretical focus, is that strictly for the Statistics degree? I was also considering the Statistics & OR combined degree, but after comparing the curriculum difference, I'm concerned the latter will have less theory and be more applied.

EDIT: Just realized you were referring to the Biostat degree. Thanks for tip. Do you have an opinion on the Stat/OR degrees though?

I have mixed feelings about this. Everyone on this forum, and elsewhere on the internet, is of the opinion that they are pretty much the same degrees. But, while my interests in statistics itself is rather broad (high dimensional data analysis/inference, ML, topological perspectives, probably more that I'm not aware of yet), I am picky in my applications. For whatever reason, I'm not particularly interested or inspired by biology as a science. I am aiming to apply research in the area of environmental science, ideally climate/weather forecasting, possibly another physical science like astronomy. I would also be interested in sociology applications, specifically regarding economic inequality, effects of public policy on such problems, etc. So, to that end, most of the programs I have picked have at least some faculty applying research to those domains.

What do you think? Are there flaws 

If you're interested in environmental, I would say biostats is actually better. There's a lot of work being done in environmental health, which includes climate change.

The STOR department as far as I know is almost completely concerned with statistical theory and in particular probability theory. There's a few people there working on ML (e.g. Liu), but there's at least as many working on ML at UNCs biostats department (e.g. Kosorok, Zeng). Biostats also has a long history in dimension reduction because for example genetic data infamously is a high dimensional problem. 

As far as sociology applications I'm not sure. I have had experience in both statistics and biostatistics departments, and I have found that the former doesn't care much for social justice research. If that's what you end up doing, that's great, but you could also just find some fun data set about, say, sports and apply it and that would be just as good to them. 

Since biostats programs are usually housed in public health schools, the applications have to be related to public health in some way. So the constraint isn't bio so much as public health. But if you want absolute freedom, maybe biostats isn't the best. Still, you won't have to do what you did for your PhD forever. 

[bayessays]

Agreed that if you like environmental applications, there's a lot of that in biostat departments. Lots of social science/public health research too. You would get a top notch theoretical education at Washington/UNC Biostat. The culture is a little different, but they'd set you up extremely well for a tech job because you'd have lots of experience working on teams on applied projects.