All Articles

Mathematics: the foundation of a dream career

Author: Lona

Jason Schissel (mathematics and physics, ‘01) grew up wanting to be a mad scientist.

“When I was six, my mom caught me at the stove standing on a bunch of pots and pans, mixing all the household chemicals,” he said.

Today, Schissel is a scientist involved in a different kind of explosion: the rapidly advancing field of big data. In 2013, he was hired as a senior data scientist at LinkedIn, the world’s largest professional network, headquartered in Mountain View, California. Founded in a living room in 2002, LinkedIn was purchased in July by Microsoft for $26.2 billion.

Schissel’s STEM (Science, Technology, Engineering and Mathematics) education in the College of Liberal Arts and Sciences gave him a solid mathematics foundation to pursue a variety of careers. Following his graduation from Iowa State, Schissel earned a Ph.D. in physics from UCLA with a dissertation on high energy astroparticle theory.

“In some ways, it was opposite of what I do now,” he said. “A big part of high energy physics is finding rare events. Big data for consumer products and monetization is mostly about finding common trends to serve a big audience, not projecting individual events.”

Schissel was working at defense contractor Raytheon when he became interested in a hot new field: data science. Networking through a friend helped him get his foot in the door at LinkedIn, and based on his strong quantitative foundation, he landed a job perfect for his math and science background.

“Friends said, ‘Oh, you went on LinkedIn and found a job somewhere?’ I would say, ‘No, I got a job at LinkedIn.’”

“Math is so flexible,” he said. “It gives you that quantitative thinking. You may not see as many jobs specifically seeking mathematicians, but it’s more the application of it.”

At LinkedIn, Schissel collaborates with a product data team on the complete product life cycle.

“When we discuss potential projects, I take recommended ideas and consider the addressable market as well as previous products and say, ‘Okay, up to 100,000 users could use idea A, but 1 million people would use idea B.’ I love the opportunity to be involved thinking about products.”

In one example, Schissel’s team worked to create an internal measurement score for a LinkedIn user’s profile to better understand what makes a profile most valuable to the user.

“The math behind the Profile Complete Score is one of the most standard tools in the industry data science toolkit: the logistic regression,” Schissel explained. “Basically, we wanted to explain an outcome, such as number of profile views a profile received, in terms of several different input factors.”

For each profile section, Schissel and his colleagues assigned either a “1” for a section that was “good enough,” or a “0” for a section that wasn’t. For example, a profile that included a photo, a summary and a position, but lacked an education section, could be represented as . They also included other inputs like user behavior to control for non-profile related factors that might impact outcome variables.

“LinkedIn has literally hundreds of millions of data points like this,” he said. “Using filtering then subsampling, I reduced down to a cleaner, more homogeneous and more manageable dataset of a few hundred thousand members. Then I used regression to assign weights to the importance of each individual field. By applying these weights, I can give every profile a ‘score’ that correlates pretty well with the full range of my output variables, which include profile views received, but also various business-centric metrics.”

Data science is still inventing itself as a field, becoming bigger and more varied, Schissel said, and artificial intelligence and machine assistance will also help shape the field’s future. For current students or alumni interested in consumer data science, a mathematics degree paired with a liberal arts education can be a great fit, along with a solid foundation in coding.

“Absolutely a strong understanding of math is important,” Schissel said. “There are many kinds of data science. What we like on the consumer science side is a strong quantitative background plus social science. I minored in philosophy and took classes in psychology. Learning how to think like a consumer is valuable.”

As an Iowa State undergraduate, Schissel could not have imagined his current job. LinkedIn, and many of today’s major internet companies, had not been invented yet.

“No one said the words ‘big data’ then,” he said.

It is proof a mathematics degree will always be a logical choice, no matter what the “mad scientists” invent next.

By Stacey Maifeld