Michigan Data Science vs. Columbia MSBA (Transfer to Columbia Data Science?)

[ansdPdms]

Hi all,  I'm deciding between Columbia's Masters of Science in Business Analytics and University of Michigan's Masters of Science in Data Science program. Since they are in different industries, I'm very conflicted. I got waitlisted from NYU and UW Data Science, got accepted to Cornell's MPS in Applied Statistics (Data Science), ORIE at Cornell Tech, and Georgetown Analytics. 

Economics major and Statistics minor at a top 3 liberal arts college, with some cs background (3 CS classes, As, and software engineering coding bootcamp) and involved in some research (data stuff, not my own research but for professors/post docs). I think my end goal is working as a data scientist at a consulting/finance firm, but I'm open to other data science roles. Not interested in PhD. I was leaning towards Michigan because of technical complexity, so I'll have a wide variety of career options, but everyone's telling me to choose Columbia because of its name value, resources, and geographical advantage (i.e. recruiting and networking). I have a week to decide - any advice/input would be appreciated!!! Thank you!

+) I am thinking of trying to transfer to Columbia or NYU Data Science; in that case, which decision is better? Transferring from Columbia's business analytics to Columbia Data Science, or UMich Data Science to Columbia Data Science? I heard Columbia DS program admits transfer students during spring semester, but I'm not sure how the program will like an applicant from business analytics, since it is not technical enough?

 

[juilletmercredi]

Michigan has great name value and resources as well. I work for a large tech company and we have more new hires from Michigan than we do from Columbia. Part of that is sheer size, but most of it is because Michigan has an excellent reputation as a research university that prepares great tech talent.

I would not enter a master's program with the intention of transferring. You should either enter a program with the intent of staying in it, OR you should wait a year and reapply. The reason is because 'transferring' at the master's level is rarely a straightforward transfer - usually, graduate programs will accept up to a semester's worth of credit from the other program, and sometimes not even that. So you spend a whole year (and $$$) on your classes and then you only use half of that or less in your new program. Plus, the business analytics program at Columbia is only three semesters, so it'd be a waste to transfer when you could just spend one more semester and be done.

The only way I would even consider this is an intra-university transfer - for example, going from Columbia's business analytics program into Columbia's data science program. The chances that you can use most of your existing coursework in the new program is much higher. (And actually, looking at it, it's really not - there's no overlap between the required first-year courses in the MSBA and the required core courses in the Columbia data science MS.) But I still wouldn't attempt it, personally - I'd say that if you 100% knew you wanted a data science program, then choose Michigan or wait another year and reapply.

I was trying to do a straight comparison between the programs, but it's difficult. At first and second glance, the Michigan program seems more technical and more of a straight/traditional data science program. But Columbia's industrial engineering/operations research department is very technical and well-respected, and a lot of the classes that have names that seem less deep in statistics actually might be quite deep and technical. I think it's probably mostly dependent upon the electives you choose to flesh out your Columbia program - you could theoretically take really technical coursework (like Bayesian Modeling & Computation, Applied Multivariate Statistics, or Data Mining for Engineers) or less so (like Managerial Negotiations, or U.S. Healthcare System, or New Product Development). Michigan doesn't have as many diverse electives - the ones you can take are all clearly focused on statistics and computer science. I think a lot of this is also dependent upon the type of internship you do over the summer, because that's where you'll learn the tools and skills to solidify the knowledge and make you more appealing to employers. But you could get a great internship from either place.