Data science is the new "Hot Major"

Yes, there are many excellent posts here. They currently hire graduates for these positions with strong knowledge of math, statistics, comp sci, and econ. If you have a major in one of those four and solid knowledge of the others, you will be a strong candidate.

My son majored in math/econ/behavioral econ as an undergraduate. He co-founded a software company with three other kids while a senior and ran it for 1.5 years. He then was admitted to study for an MS in Data Science/Computational and Mathematical Engineering and an MBA at a very strong school. All of his classmates in DS were very bright kids who immediately walked into jobs in Silicon Valley (maybe one stayed for a PhD). He thinks it is a great degree. I have the sense that math skills are already very important and as other people said, will likely become more so.

My son doesn’t intend to be a data scientist (in part because he doesn’t love the coding and in part because he likes being an entrepreneur) but expects to hire data scientists.

Although I’m thinking this post belongs on the “graduate school” forum, most here on this thread seem more likely to have better input. D has just recently decided that she wants to continue on to at least a MS if not PhD. She is currently a junior studying in Budapest in a SoftwareEngineer/CS program. She is a double major Math/CS and has nearly a 4.0 (I think its 3.96 or something). Her study abroad program is kicking her butt and she has finally met her match which has been humbling…but I digress. FTR, her study abroad does not affect her GPA but it has given her some insight as to her strengths and a better global view of her peers. I mention this only to highlight that Math is her real strength.

She is interested in AI/ML. We’re looking at programs but its a learning curve…what school, what program, how does financial aid work. So lets start with the basic, and this is where I need your help…for a student with great Math skills and decent/good CS skills hoping to work in some sort of AI/ML capacity, what type of advanced degree should she be looking at? Math or CS ? And what schools should we be researching? I can’t believe we’re back at starting point again lol.

Thanks everyone for your help.

@NEPatsGirl, AI/ML is usually going to be taught in CS departments. Having a strong math & statistics background is very helpful, though. There are so many schools where you can study the subject now that I wouldn’t even know where to begin with suggestions on where to go.

When I got my Masters, I went to night school and my employer paid for it. Otherwise, I have no idea how financial aid works for graduate school.

Thank you @simba9. I’m not sure she has much background in stats but at least a couple of classes. Unfortunately, we’ve discussed going to work for a company that will pay for advanced degree but she isn’t buying it lol.

The beginning of the list is UmassAmherst, Virginia Tech, possibly BU. But it will all come down to finances in the end, just like undergrad. She is open to anyplace, literally anyplace, but would like to live in a city if possible. Of course, UMA and VT don’t fit that bill. She will be taking the GRE exams in the fall.

@NEPatsGirl

If one does choose to go directly into AI/ML, I would pay attention to how much research into the field is being done there. This both often highly correlates with the numbers of masters classes in the area and the opportunities for research at these schools.

http://csrankings.org/#/fromyear/2007/toyear/2018/index?ai&mlmining

CMU, UT Austin, Georgia Tech, Rutgers, UMD-CP, and USC would be some top options in cities.

For advanced AI: CMU, Stanford, MIT, Cal

1 Like

I’d think any of the Big 10 schools would work for graduate-level AI/ML work. In fact, just about any big school of say, 20,000 or more and a graduate CS program, would work.

Old “news.”

UT Austin and Georgia Tech are on the list, just not sure she is a “likely” at these caliber schools. Most of the websites are saying 3.0 or better and name a few prerequisite classes, all of which she has had. Given the difficulty of getting into a top 50 undergrad school these days, what is the possibility of a good student getting into places like CMU? Its pretty clear that Stanford, MIT and Cal are not within reach.

@epiphany, please explain.

I actually think it’s easier to get into graduate school for CS at places like Stanford and CMU than it is to get into the undergraduate programs. That’s because they have graduate programs for working professionals that they view as cash cows, so they’ll accept as many students as they can before it starts diluting the brand. (“Easier” doesn’t mean easy, though.)

Anyway, as is so often mentioned, ranking of your school doesn’t matter that much for CS. If a graduate of any reputable school can show they know their stuff, employers would be very happy to hire them.

@doschicos- Tufts just implemented Data Science as a new major within Engineering - as a joint offering between the Electrical and Computer Engineering Department and the Computer Science Department. All the courses already existed, but that may not be true at all schools.

@juillet - I don’t think it is a fad. The driving force for the new major was a CS professor who also does data science work outside of Tufts who could not find people to hire with the right skill set. Note that it is already possible to program a computer to recognize patterns in data. It is a specialty within AI called Machine Learning and it is one of the hottest fields within CS.

As implemented at Tufts, Data Science is an Engineering degree (i.e. 38 courses, including: 6 Humanities,/Arts/Social Science, 1 Ethics, 28 STEM and 3 free electives). It looks like an ABET accredited CS degree with electives in Machine Learning, Secure/Networked Databases, Human Computer Interfacing (HCI) and Computational Data Science Theory. There is also a set of Applied Computational Data Science courses that are tied to an area of interest and a related 2 semester Capstone Project. In theory a student could have put together a similar program on their own, but they would not have had the experience/expertise. I have included links to information for those who want to try to design their own.

@hebegebe - I agree that it can be an interesting field

http://now.tufts.edu/articles/tufts-offers-new-data-science-degree
https://engineering.tufts.edu/cs/bachelor-science-data-science
https://engineering.tufts.edu/cs/sites/default/files/BSDS_DRAFT_DegreeSheet.pdf
http://www.cs.tufts.edu/t/courses/description/undergraduate
http://math.tufts.edu/documents/coursedescfall2014.pdf

@NEPatsGirl

Look at Reply #1 :slight_smile:

@Mastadon

To be clear, I don’t mean that I think data science - as a field, a discipline, or a way of doing analysis - is a fad. That’s here to stay. I work in tech and we have more data than we know what to do with; we need AI/ML specialists and data science to continue to analyze the data. In fact, as a behavioral scientist that works in the field I work closely with data scientists and we need their insights to paint the pictures of behaviors; the methods that I use (even quantitatively) can’t handle the volume of transactions on the Internet or in the cloud these days.

I mean that the enthusiasm for the data science major amongst college-bound students is probably a fad, as well as the universities/colleges building these programs. Many students won’t be interested in the intricacies of computer science and math/stats required to major in data science (much like probably most of pre-med majors change their minds). And building a data science program of quality is actually difficult and expensive. Colleges already have some level of difficulty finding and retaining computer science instructors, since people with advanced degrees in CS can make so much more money doing just about anything else (especially if they have research and development expertise, and especially if they know enough about AI/ML and data science to teach it!). My guess is that many top universities will have a minor or certificate in it, some will have the major, and at the majority of schools the closest you’ll get is a double major in math/stats and computer science or an independently designed major.

Besides, whatever the ‘hot majors’ are changes every 5-10 years anyway…I mean, data science didn’t exist as a major 10 years ago, and 10 years from now there will be a bunch of new jobs and fields that don’t exist today.

With the increase in popularity in these type of degrees parents and students need to pay attention to the programs they investigate. It’s one thing to market college programs with heavy data science, AI, and ML it’s another to have the professors needed. @juillet is spot-on in cautioning about the ability of colleges to attract and retain expertise in these areas. Take a lok at this NY Times article: “A.I. Researchers Are Making More Than $1 Million, Even at a Nonprofit” https://www.nytimes.com/2018/04/19/technology/artificial-intelligence-salaries-openai.html

Computer science and software engineering are very faddish fields. Data Science/Machine Learning/AI won’t go away, but the hype will eventually die down and they’ll attain sub-field status within CS, akin to databases, networks, and graphics.

One of the things I enjoyed least about my career as a programmer was constantly having to learn the latest industry fad, knowing full well that it would be replaced by another fad in three years. Often the latter fad was no better than the previous one, but people always assume that whatever comes after is better than what came before, so you have no choice but to learn them if you want to remain employable. It always seemed like most of my time was spent learning things rather than creating things. (That’s my little rant about software fads.)

The bulk of what programmers use on the job is self-taught, so while people will go to school to learn fundamental CS and programming concepts, and basic math and statistics, most of the practical uses of Data Science/Machine Learning/AI will be learned on the employee’s own time via books and online tutorials. So I’m not sure how critical Data Science or Machine Learning programs in schools will end up being.

Tufts just published a guide to the new major. It is sort of interesting. Apparently, they have no plans to offer a version of the major outside of the School of Engineering (SOE), which is much smaller than the School of Arts and Sciences (A&S), - due to costs. Tufts has about 5 or 6 professors who teach courses in AI/Machine Learning

Can’t provide the link because apparently it is a Google Doc…

As a Recruiter in the SF Bay Area, this is a hot topic right now. Data Science roles tend to fall into two ‘silos.’ Data Engineering is the role that collects, and structures large data (creates order from chaos). Data Scientists use the tools to analyze the data, create meaningful information from that data. Both roles are CRITICAL and can’t work without each other.

The unicorns are those that do both - they are rare and very expensive. The funny thing is that most jobs are posted as if these unicorns are as common as a zebra in a zoo

Aren’t most jobs posted with a wish list of “requirements” that are very unlikely to be met by any applicant, and don’t most job seekers realize that and apply to all jobs that they meet even the smallest subset of “requirements” for?