Varied experience builds strong data science background
With the demand for data scientists still booming, many are flocking to the career, and they bring various skills with them to the field of data science. Data science can also be applied in many different fields, so often those without a title related to data science use these skills in their everyday work.
Skills gained in varying areas, from mathematics to cancer research to astrophysics, can be valuable in a data science career.
Where do data scientists come from?
Especially since colleges and universities are only just beginning to offer data science majors and master’s degrees, few people entering the field have a specific data science background.
One common background for data scientists is mathematics. Speaking at the Women in Data Science conference in Cambridge, Mass. on March 2, Moon Duchin, associate professor of mathematics at Tufts University, described her background, which mainly involves using data science in voter redistricting.
“It’s all math skills that you build [and those skills] help you better when they improve,” she said when asked how her background in mathematics helps her data projects.
Duchin uses math and data science for the purpose of civil rights. She currently works with city and state officials to redraw voting districts so groups that are currently underrepresented can have their votes counted more accurately.
Building expertise in mathematics is crucial in data science, said Duchin. Mathematics helps “push the frontiers in machine learning — ask why and how” models work.
What brings people to data science?
But it’s not only a math background that prepares a person for a career in data science. Katie Montgomery, who attended the conference, is new to the data science community. With a background in communications, Montgomery is a staff assistant in the capital giving team for the Faculty of Arts & Science at Harvard University. Data plays a major role in increasing fundraising capital. Her team works with data to determine where to focus fundraising efforts.
“There must be a system other than humans to pull that information,” Montgomery said. “I’m here to see that there are other ways to get there without a background in data science.”
With the rise of embedded BI platforms and the increase in citizen data scientists, it’s increasingly common to have more people in similar situations.
Vasudha Shivamoggi, senior data scientist at Rapid7, didn’t come from a true data science background. She earned a PhD in physics and later moved into a career in data science.
“Long story short, I needed a job, and I moved into data science,” she said in an interview at the conference about how she decided to make the switch from physics. “I think physics has a lot in common with data science already.”
To Shivamoggi, there’s almost a straight line between a quantitative science like physics and data science.
“A lot of the focus is on modeling so it’s quantitative, it’s mathematical,” she said, “but also we know what we produce inherently is an approximation. And that’s true in physics as well. We don’t ever get the full picture, but we’re trying to figure out what’s the simplest model that will still answer our questions.”
Data science: Just like research?
Jess Stauth, managing director at Fidelity Labs, said in a presentation at the conference that a history of scientific research could help data scientists excel in their roles.
Stauth pointed out that theoretical or research-based sciences and applied sciences, such as data science, are often viewed as two entirely separate things.
“These are two different things, but I think it’s a disservice to separate them so strongly,” Stauth said. She added that there’s a strong value in teams with diverse skills working together. A person with a more general science background can bring valuable skills to this kind of team.
Stauth has worked in data science teams in startup companies. She said that in a startup atmosphere data scientists are asked to wear many different hats and sometimes step into several roles. That’s why having a diverse data science background can be an advantage for people pursuing the discipline as a career.
“A data scientist can also be an analyst or a business user — those don’t have to be separate,” Stauth said. “They’re two different hats that you have to wear.”