By Katie Kackenmeister
When talking about data, bigger doesn’t always mean better.
“We’ve reached a point where getting more of the same kinds of data isn’t as helpful as it used to be,” says Daniel P. Robinson, an associate professor of industrial and systems engineering who specializes in mathematical optimization and data science. “For some applications and problems, even if we had much more data, we wouldn’t see a significant benefit.”
Where there is room for improvement, he says, lies in developing new scientific approaches and tools to handle that data—and educating more people with the specialized mathematical knowledge and computational skills to make sense of it all.
Robinson is one of three Rossin College faculty co-founders of a new cross-departmental master’s degree program focused on training the next generation of data scientists to meet that need.
Accelerating opportunities
“If you want to be on the cutting edge of almost any field, you need computational and data science expertise,” says Brian D. Davison, a professor of computer science and engineering. “If you can exploit your data in better, more intelligent ways—whether you’re a microbiologist or a civil engineer—you’re going to push your field further.”
Davison directs Lehigh’s undergraduate minor in data science and has taught the intro course in the discipline for the past six years. Recently, he’s worked with Robinson and Parv Venkitasubramaniam, an associate professor of electrical and computer engineering, to design a data science master’s degree program that’s interdisciplinary from both faculty and student perspectives.
Lehigh’s new MS in Data Science program brings together professors from computer science and engineering, electrical and computer engineering, and industrial and systems engineering to teach graduate students the fundamentals of data science from varying viewpoints.
On the flip side, the approach will allow students from a wide range of backgrounds to gain the qualifications necessary to tap into the wealth of data science jobs, many of which require an advanced degree. (The U.S. Bureau of Labor Statistics projects jobs in the field to grow by more than 30 percent between 2020 and 2030; the current median salary for a data scientist is nearly $100,000.)
The program, which launches this summer, “will provide a new opportunity for people who want to develop data science skills, even if they don’t have a background in computer science or statistics, to get this valuable, in-demand credential,” says Davison.
Leveraging ‘domain expertise’
As Davison sees it, data science is a broad concept, encompassing methods of collecting and extracting data, processing and validating it, and putting it to a practical use, whether that’s a graphic visualization of the spread of COVID-19, for example, or a prediction of a prospective borrower’s likelihood to default on a mortgage generated through machine learning.
There’s an ethical component, too, Robinson adds: “How do we understand and use data in a responsible way, and apply and improve algorithms to avoid bias?”
And many of the techniques used in the field have roots in sensing, networking, and signal processing, where Venkitasubramaniam’s interests lie.
Prospective students need a strong foundation in math (like calculus and linear algebra) and basic training in computer science (typical for many undergrad STEM degrees and some business-related majors), but no prior data science coursework is required.
Classes will be taught in a hybrid (both in-person and remote) format to accommodate full-time students—including those who desire to complete the master’s degree over 12 months—as well as professionals who want to study part time.
“It’s very valuable to have some ‘domain expertise’ in the area where you are applying computational and statistical or mathematical methods,” says Davison. That’s one reason why in the real world, he says, data science is often conducted in teams. “If you don’t have that expertise, you can create a great system that is of no value, because you don’t know what is valuable in the particular problem you are trying to address.”
Tech giants like Google and Amazon typically come to mind as companies that employ data scientists, Robinson says, but as sectors like health care, finance, manufacturing, and even nonprofits in the “data for good” space latch on to data analytics, the career pathways—and opportunities for social impact—are expanding.
For example, he says, in medicine, the ability to combine data science with artificial intelligence tools can help improve the accuracy of diagnoses and predict health outcomes to better guide treatment decisions.
“If you’re interested in making an impact on people, I can’t think of a better way to do it, broadly speaking, than through data science,” he says.
Looking down the road
Core coursework for the MS in Data Science covers topics such as algorithms, mathematical optimization, statistical modeling, advanced computing for machine learning, and ethics. Students will gain practical experience developing and applying methods to extract relevant information from data. Elective courses allow for specialization in areas that match students’ interests or current professional focus.
Robinson is particularly excited about the opportunity for master’s students to engage in projects with computationally focused faculty, including those affiliated with Lehigh’s Institute for Data, Intelligent Systems, and Computation (I-DISC).
“Getting involved in a research project can be life altering,” he says. “It can change the course of a career.”
As the program grows, the team envisions collaborations with faculty from Lehigh’s other colleges.
“The buzz around data science isn’t just about the impact it’s had; it’s also about potential,” Robinson says. “What it can do hasn’t been fully realized.”
-Photos by Christa Neu
