[Fall 2018 Statistics Applicant Thread]

10 hours ago, biostatboi said:

@aaaapvawgw they removed that program from the online application. I received an email in October that said:

"We have decided to remove that joint program option from the on-line application.

Instead, students interested in the joint Ph.D. degree in Statistics and Machine Learning should apply to the PH.D. program that best aligns with their research interests (PH.D. in Statistics or Ph.D. in Machine Learning)."

This is also said here: http://www.stat.cmu.edu/phd/statml

When did you submit your application?

I'm a little confused. Does this mean people who expressed interest in ML in their SOP's are automatically considered or or that the process/application for joining begins once you are already admitted/attending one of the stats or ML phd? I'll also ask them via email.

[Statistics PhD Profile Evaluation/ Advice on what to prioritize]

On 11/24/2017 at 9:54 AM, statsphdapps said:

Update: I wound up scoring and 820 on the Math Subject GRE (82 percentile)! Given my strong GPA and math courses from a well-known undergrad institution, would it help or hurt me to send such a score to schools that do not require the subject GRE but strongly recommend it such as Chicago and Washington (I feel like I'm in a grey area as to whether or not sending the score will be helpful). What about for schools that say the test is optional, and you can report scores but they don't strongly recommend it?)

 

I got the same score and would also like to know the answer to this question.

[Prof won't let me take Grad Mathematical Stats]

Thanks for the response Innominate. I guess it isn't as big of a deal as I thought. 

On 4/9/2017 at 3:32 AM, Innominate said:

you can definitely take something like data structures or numerical linear algebra, etc. which aren't technically stats but will boost your breadth.

Yeah, I am a pure math major and CS minor who has mostly taken classes in pure math. I actually don't have a lot of programming experience (only 2-3 classes) since I am more interested in applying to grad schools on the theory side. I have like no background in R or the similar applied stats languages. I'm hoping this isn't that big of a deal, but I think I'll take a class or two on programming-related things if I can.

[Prof won't let me take Grad Mathematical Stats]

Hello,

I'm a junior who plans to apply to stats phd programs next fall. I took the undergrad mathematical stats at my school and wanted to take the grad version (since there is only 1 semester of undergrad math stats) in the fall, but the professor refuses to allow me to take it.

The only thing he asked for was my GPA (~3.75) and coursework and, in my humble opinion, his decision makes very little sense. I got an A in the undergrad math stats course and have also taken grad real analysis, complex analysis, topology this year, all with A's. I mentioned all of this to him and he said my grades in calc III/IV (got B+'s freshman year) are a concern and that I should instead take "easier" courses to get a better GPA if I want to apply to grad school.

It's also not the case that the grad math stats course is somehow significantly harder than the undergrad version. They both use the same book (Casella & Berger), so it's overall a bit strange that the prof believes I am totally unprepared for it. In any case, I don't want to argue with him, so I decided to just not take the course.

How bad do you think this will affect my application? It looks like I'll only be applying with 1 semester of mathematical stats now. Is it fine if I just stick to pure math/CS courses for the fall? My school doesn't really have a dedicated stats dept so there aren't really many options among stats courses.

 

[Importance of Stats Research for PhD programs]

I'm a junior undergrad with a mostly pure math background. However, very recently, I'm becoming more and more interested in stats through self-learning and will probably apply to PhD programs in it. However, outside of the standard undergrad probability theory, data analysis, operations research, and mathematics statistics classes, I don't have any actual research experience working with datasets or doing anything stats-related. My only hands-on coding project was using Python for a pure math research project last summer.

I've been accepted to a pretty decent pure math REU (lots of submitted and published papers among its participants), but I'm wondering if it might be better to look for a stats-related research opportunity instead (or do something like SIBS or RUSIS).

Also, I'm gonna try to take the grad mathematical statistics course at my school in the fall, but I'm wondering what else I can do in the coming year to make my application more attractive? I noticed a lot of the phd students at top programs have work experience coding at startups/tech firms and/or applied stats REUS under their belt.