Other Master's Degree Biostats masters vs Math B.S. for medical data science

[Teufelhunden]

So, I'm convinced that medical data science is going to be a thing someday (great article here, btw), and I'm trying to figure out how to transition into it. I just completed the first term of Udacity's Data Analyst Nanodegree, and I'm quickly figuring out that math/stats is really at the foundation of all this (where Python, R, SQL, etc are its tools, and data wrangling, EDA, etc are its processes), and so I've decided to pursue formal education in math. I'm in Cleveland, and CWRU has a Biostats M.S. that I'm considering applying for. My other option is to simply go back to school and chip away at a second bachelors in math. The latter plan would be more flexible, and I think would give me a more solid math foundation (the biostats masters seems more like 'applied' math to me, and relies heavily on SAS). The biostats masters, on the other hand, is a graduated degree from a somewhat prestigious university and would probably help open more doors in the future (and allow me to network with folks in Case's Department of Population and Quantitative Health Sciences). I guess my real dilemma is this: I feel that with a math undergrad degree, I'll have a solid math background that'll prepare me for any applied math discipline, whether that be biostats, healthcare data analysis, healthcare informatics, and even the data science subfields (machine learning, deep learning, AI). I realize I haven't asked a cogent question, but rather shared what's been bouncing around my head for the past few months while trying to decide which path to pursue. So, I thought I'd bring this to my esteemed colleagues here...knowing there's probably a lot of folks here with backgrounds spanning all these disciplines who might be able to offer some insights. Thanks!

---

Members don't see this ad.

 

[751866]

So, I'm convinced that medical data science is going to be a thing someday (great article here, btw), and I'm trying to figure out how to transition into it. I just completed the first term of Udacity's Data Analyst Nanodegree, and I'm quickly figuring out that math/stats is really at the foundation of all this (where Python, R, SQL, etc are its tools, and data wrangling, EDA, etc are its processes), and so I've decided to pursue formal education in math. I'm in Cleveland, and CWRU has a Biostats M.S. that I'm considering applying for. My other option is to simply go back to school and chip away at a second bachelors in math. The latter plan would be more flexible, and I think would give me a more solid math foundation (the biostats masters seems more like 'applied' math to me, and relies heavily on SAS). The biostats masters, on the other hand, is a graduated degree from a somewhat prestigious university and would probably help open more doors in the future (and allow me to network with folks in Case's Department of Population and Quantitative Health Sciences). I guess my real dilemma is this: I feel that with a math undergrad degree, I'll have a solid math background that'll prepare me for any applied math discipline, whether that be biostats, healthcare data analysis, healthcare informatics, and even the data science subfields (machine learning, deep learning, AI). I realize I haven't asked a cogent question, but rather shared what's been bouncing around my head for the past few months while trying to decide which path to pursue. So, I thought I'd bring this to my esteemed colleagues here...knowing there's probably a lot of folks here with backgrounds spanning all these disciplines who might be able to offer some insights. Thanks!

You should consider an MS in statistics also, because you can still take courses relevant to biomed focus. "Data science" and those other hot topics are just derivatives, often watered down, of Statistics as a whole. Machine learning is nothing new, contrary to new articles and the "field" of machine learning. Get a real statistics/biostats background and you'll be more equipped than any of those other fields. Get a healthy exposure to coding and you'll be on par with "machine learning" people in that regard and your foundational knowledge will generally be much better. Good on you for pursuing the mathematics to get a good degree.

Edit: Sorry, I thought you meant you were going back to school for a mathematics bachelor degree before pursuing the stats/biostats masters. If you only have one choice, go for the stats or biostats masters.

 

[wholeheartedly]

biostats or just stats will probably position you best and in the most efficient way possible. If you do pure math, it'll still give you a great foundation, but with a lot of extra stuff you won't need and missing some stuff you will need. the application part of these things is tricky, so that's actually a pretty important part of the education.

keep an eye on bio informatics as well.

 

[n3t]

In my opinion, a master’s degree is more marketable. If you are not interested in SAS, it may be good to look at some other graduate schools, as different schools have different programming focuses. Some schools put an emphasis on SAS, while other schools put an emphasis on different programs. You will likely gain exposure to multiple programs, but if you believe something different is more important to your field, it cannot hurt to look at other schools as well.

 

[Excision]

It's already a thing in healthcare, although a bit behind the curve of other industries.

Organizations and consulting firms are becoming increasingly metrics driven to find competitive advantages (e.g. where do we place that urgent care?). It's a pretty cool field. I'm currently learning sql, excel vba, and reporting tools like power bi and tableau in an effort to develop a business intelligence skill set.

 

[PubPhD]

I'd skip any 'bio' emphasis on math. More often than not, these courses skip the theoretical/proof based understanding of the methods you're using so you won't have the core foundation needed to excel in the field (and I've seen some awful mistakes made by those who plug'n play with methods). You pretty much only need equations, sql, python (and the respective ML/numpy/scipy libs), and a good understanding of basic CS to avoid any horrible implementations to get a job.

Having worked for one of the largest health ML corporations, I'd take the 2nd bachelors. MS's gives you a hammer, but everything will look like a nail. Spending time to get a good foundational understanding of the fields you'll be working in (measure theory, algebra, etc) will let you have a solid intuition (from first principles) behind what you're doing.