What to self-study before starting a stat/biostat PhD program

[sisyphus1]

Looking for advice/ideas on what to self-study before starting a stat/biostat PhD program (with the hope that I will get into at least one program in the coming weeks!).

 

Some potential ideas:

 

- Get a better biology background--I have practically 0 exposure to biology, and am thinking of reading up on genetics/neuroscience textbooks. I have an interest in applying machine learning techinques to biostatical problems, and I'm guessing that it can't hurt to know more about biology in a biostats program.

 

- Learn (like really learn) a programming language and/or a statistical package. I've had decent exposure to MATLAB, SAS, R etc., but I wouldn't say I am good at any. I am thinking of getting a really good foundation in R/Python so I can be of immediate use in (say) a research setting.

 

- Learn some graduate-level math--specifically PDEs and Fourier transforms. I think I have the basic prerequisites for a stat/biostat program (analysis, linear algebra etc.--I was a math major), but I've seen some really cool papers where people were able to marry mathematical concepts from different domains (e.g. there was a guy from CMU who expressed distributions over a combinatorically large solution set in terms of Fourier transforms. I didn't fully understand it but thought it was pretty neat). Can't hurt to know more math.

 

Obviously this may depend on the type of program I (knock on wood) get into (theory vs applied, stat vs biostat). But any thoughts/suggestions are welcome.

[cyberwulf]

I think you will derive the greatest benefit out of becoming more familiar with R. It's something you can learn quickly in a short time (unlike advanced math topics) and will be invaluable in many courses and RA projects (unlike specific biological knowledge).

[superbean]

Coursera just started a four week long course on R. It is free, and I'm taking it now too. There are downloadable videos, some quizzes and some homework assignments

https://www.coursera.org/course/compdata

Edited January 3, 2013 by superbean

[wine in coffee cups]

Second learning R, being more familiar with it would have saved me a lot of time. I also really wish I had brushed up on calculus since it has been 9 years since I last did integration by parts and polar transforms and such. LaTeX is good to pick up in advance if you are not already comfortable. All the fancier math that is somewhat more recent to me has not been of much use thus far.

[ANDS!]

- Get a better biology background--I have practically 0 exposure to biology, and am thinking of reading up on genetics/neuroscience textbooks. I have an interest in applying machine learning techinques to biostatical problems, and I'm guessing that it can't hurt to know more about biology in a biostats program.

Don't bother. There will be people to dumb down the science to a level you need to know it - if it is even neccessary (likely to not be).

 

- Learn (like really learn) a programming language and/or a statistical package. I've had decent exposure to MATLAB, SAS, R etc., but I wouldn't say I am good at any. I am thinking of getting a really good foundation in R/Python so I can be of immediate use in (say) a research setting.

There will most likely be a class for this. If you have an introduction to the programs (to I would say a point that if you don't know how to do something you can Google it effectively) you are most likely to be above what most of the students will have.

 

 

- Learn some graduate-level math--specifically PDEs and Fourier transforms

Sounds more undergraduate level (though upper division) to me. But I would say it's overkill for what you are likely to encounter your first year. A lot of the more advanced topics you'll pick up as you get more familiar with what it is you actually want to do.

I would say if you want to maximize utility, get good with R and SAS (if you don't envision taking it next year - free credits though); to me that would give the most benefit (internships generally are looking for people with facility in this).