CMU MSP Vs. Columbia's MA in QMSS Vs. Cornell's MPS

[kevy]

Before we start, a little about me:

 

Goals: After getting my master's, I want to work for a couple of years to build experience and to pay off some loans. My end goals is to get a phd in quantitative psychology and become a professor.

 

What I am interested in: I am interested in data science but I also want to learn more about the statistical side of it.

 

Right now, I am stuck between choosing relevance (Columbia's QMSS program) and practicality (CMU and Cornell). While it may seem clear to just pick Columbia, I am actually a little worried about the quality of the curriculum. For some reason, it just seems.... not as complete for the data science track. I am also a little scared about building a foundation from only the social sciences point of view. On the other hand, CMU's MSP seems like a better fit for the short run and CMU is pretty well-known for machine learning. However, the main focus for CMU's MSP program is applied stats. But after speaking with CMU's director, it seems like they will tailor the program to fit the interest of the student. On the other hand, Cornell's MSP program has a data science focus.

 

There doesn't seem to be much information about CMU's MSP or Columbia's QMSS. Any insight would be greatly appreciated. I guess in the end, I just want a program that will strenghten my statistical background while focusing on data science/machine learning. It would be a plus if that program can also prepare me for either a career in the private industry or in academia.

[edwardmen1]

Did u get accepted by QMSS? Then go for it

[kewz]

I'm taking my second class with the MSP students at CMU right now. They have a set of master-level classes designed specifically for them and also take special topics classes that are cross-listed as advanced undergraduate as well as master-level, like the one I just mentioned. As far as I know, the program is very applied - which is unsurprising, because that is exactly the point of the MSP program. This year, many of them seem to be working with a renowned cognitive scientist in the psychology department on analyzing fMRI data. There also seems to be a lot of programming/ computing, because I've heard a lot of R discussion from them in the lounge.. In general, statistical computing is pretty big at this department; after all, several (or perhaps many) professors also have appointments in the machine learning department and the computational biology department. Several phd-level core courses are also cross-listed in both the stats and the ML depts. If you are into ML, it is uncommon for undergrads across the CMU campus (i.e. not necessarily CS majors) to take master- and/or phd-level machine learning classes (10-601/10-701). I'm sure many master students from non-CS fields take them as well. The MSP director is a really nice guy; you should talk to him about such options. 

 

Just want to add a remark on data science: you don't hear much (actually, not at all) about the term 'data science' here at CMU. The philosophy here is that statistics is data science. 

 

http://www.huffingtonpost.com/american-statistical-association/statistical-thinking-the-bedrock-of-data-science_b_3651121.html

Edited April 2, 2014 by kewz