Eval: Statistics MS and or PhD (Machine Learning)

[Cauchy_Way]

Here are my stats:

Undergrad Institution: University of Colorado, Boulder
Major: Mathematics, Computer Science
GPA: 3.81

Type of Student: White Male

Upper Division Courses:

Math: Calc III (A-), Analysis I (A), Analysis II (A), Intro Linear Algebra (A), ODE (A), PDE (A), Operations Research (A), Numerical Analysis I (A), Probability Theory (A), Statistical Theory (A), Markov, Queues, Monte Carlo Simulation (A)

Next Semester: Stochastic Processes (Grad Level)

 

CS: Algorithms (A-), Artificial Intelligence (A), Machine Learning (A)

 

Others: Generally A- and As in the basic maths and computer science courses. Two Bs (one in Calc II), several B+ (all non-related classes), a B- in a history class. 

 

GRE: 168Q, 165V, 4

 

Research Experience: year long research project with two professors in math department, undergraduate thesis in machine learning area 

Awards/Honors/Recognitions: Dean's List, several (recent) scholarships

Pertinent Activities or Jobs: BB bank quant summer analyst, Data Science team, active with elementary/middle school volunteer programs

 

Letters of Recommendation:

-Two mathematics professors (research with both)

-Computer science professors (thesis advisor)

 

Plan to apply to (mix of MS/Phd Programs):

Chicago, Berkeley, Harvard, MIT (CS only), Penn, Yale, CMU, Wisconsin, Northwestern, Toronto, Cornell, Texas (CS only), UCLA, Virginia, and a few  "safes"... 

 

Am I reaching too much for some of these schools? I would rather not settle since I have an offer from the same BB bank I interned at, but I'd also prefer to further my education. Would working a year or two help at all? 

[StatsG0d]

I would say you've got a good chance to get into a good PhD program. I would maybe add on a few of the larger state schools like Penn State, Purdue, and Texas A&M (all of which have machine learning faculty in their stats departments). The Ivies are pretty tough to break into, especially for a PhD. So if you're interested in a PhD, definitely take off a few of those and add on some more realistic schools. On the other hand, I feel Virginia might be a little "too safe". As everyone says, your advisor matters more than ranking, but I still think it helps. Just do some research on good programs that have faculty that do machine learning.

For the MS, I would say you have a very good shot at getting into any of those schools.

I don't think your work experience will help unless you're doing some type of research (perhaps in applied statistics).

[GoPackGo89]

10 hours ago, footballman2399 said:

... The Ivies are pretty tough to break into, especially for a PhD
 

Are incoming classes at ivy league PhD's generally smaller than most schools? Are they more selective than say schools like Berkeley, Washington, UNC?

[StatsG0d]

On 11/30/2016 at 9:00 PM, statbiostat2017 said:

Are incoming classes at ivy league PhD's generally smaller than most schools? Are they more selective than say schools like Berkeley, Washington, UNC?

It depends. Yale has a very small program but Columbia's is quite sizable. More than anything, I would say the Ivies are heavily international in terms of their PhD student body. It's difficult to compete with people going to Peking University in China when you went to a state school in the US. 

Not too sure regarding your second question. I'd imagine Berkeley and Washington are similar because they are very top tier programs. UNC I doubt is anywhere near as competitive. 

[Cauchy_Way]

On 12/6/2016 at 6:14 AM, footballman2399 said:

It depends. Yale has a very small program but Columbia's is quite sizable. More than anything, I would say the Ivies are heavily international in terms of their PhD student body. It's difficult to compete with people going to Peking University in China when you went to a state school in the US. 

Not too sure regarding your second question. I'd imagine Berkeley and Washington are similar because they are very top tier programs. UNC I doubt is anywhere near as competitive. 

Interesting about the international competition. Care to elaborate on why students form say, Peking, are more qualified? I know its arguably the best school in China, but globally there are plenty of "state" schools within the U.S. that rank higher, given you believe the rankings are calculated without bias. 

[StatsG0d]

On 12/9/2016 at 1:45 AM, Cauchy_Way said:

Interesting about the international competition. Care to elaborate on why students form say, Peking, are more qualified? I know its arguably the best school in China, but globally there are plenty of "state" schools within the U.S. that rank higher, given you believe the rankings are calculated without bias. 

Don't get me wrong there's probably loads of bias in the rankings. But they're still good as a baseline to compare the tier of different schools.

I wouldn't necessarily say "more qualified" but I would argue they are better prepared. Many of these students will have already seen a lot of the material you will learn in your first year and maybe your second. Students that come from these schools have a good track record of being successful in the programs. 

[Residuals]

Your record should be no problem for MS programs at all these schools. You don't state which schools are MS vs. PhD so it is hard to be more specific. CU Boulder is a respectable math department so your courses will be taken seriously. Your math grades are solid except for a blemish with Calc II. Your GRE is solid. You might have difficulty at the very top schools for PhD going against students who have taken the GRE math subject test and have done well. But, being a domestic male is an advantage. Math at top state schools is not any different than the international schools. Domestic students take graduate probability at the undergraduate level so domestic students will experience courses in the graduate program early just as many international students. The substantial difference between domestic and international applicants is the test scoring. International students work intensely on problems repetitively resulting in the ability to master GRE and GRE subject tests. You have international students who have terrible English skills who score 90% on GRE English! It is a bit of a joke. But, as a result, international students have created a problem for themselves in that their scores are "discounted" relative to a domestic student. Personally, I don't view the Ivies as representing the stronger statistics departments.  The academic placements of Harvard and Columbia are much weaker than many other departments. Penn fairs okay.

[StatsG0d]

For the record, I never meant to say that I believe the Ivies departments are "better." But I do believe they are much more difficult to get into. In addition to the reasons listed above, the endowments at these schools are so huge that they don't need to rely on Federal grant funding, so that enables them to basically take the best and brightest from the applicant pool. State school programs usually do have to rely on federal funding, giving domestic students an edge. 

I'm not trying to say that your profile isn't strong or that I don't think you can get into a good program. I'm only saying the Ivies are very, very difficult to break into. For example, I got outright rejected from Columbia and Yale despite being admitted to programs better than both of those (and receiving fellowship offers from 3 top state school programs). 

[Robbentheking]

I just want to add that I have a friend from China who's told me that the dynamic in China is very different from the US. In his experience, undergrad comes with an insane amount of competition, while grad school isn't really that crazy because most of the smart kids go to America to continue their studies. 

Those university rankings are based on a lot of factors and I think it's safe to say better university ranking does not imply math undergrads know their material better. Some of these Chinese unis seem like good counterexamples.