Daniel Gillion’s course teaches students without a coding background how to apply models to a wide range of problems across political science, economics, and sociology.
3 min. read
Daniel Gillion and his class discuss current events for a solid grounding in the issues in his course, Machine Learning in Political Science, Sociology, and Economics.
The class filled with the clatter-click tapping of laptop keys, the cracking of knuckles, and the echoes of small coughs—the sounds of a pop quiz as students worked to determine which voters would be likely ones for a select candidate.
Students in Daniel Gillion’s Machine Learning in Political Science, Sociology, and Economics course don’t just discuss data analysis and artificial intelligence in the abstract. They put it into action, applying it to real-world social science problems. They are learning how machine learning models work and even how to build chatbots—no special coding or computer science background required.
“The expectation is that you come in knowing nothing,” Gillion says. “The reason why AI is just ballooning into something that’s amazing, for so many different fields, is its ability to be applied to almost anything.”
That accessibility was a draw for Erica Jiang, a second-year politics, philosophy and economics major from New York City. “Professor Gillion was willing to take us from zero to 100,” she says. “There’s not too much of a learning curve. I find it really cool that, even though I don’t know all the technical material, I can still make these working models.”
Students explore applications from finance and politics to sociology and even nursing, using machine learning to make predictive decisions about the impact of phenomena or forces. Traditionally, Gillion says, fields such as political science and economics have relied on methods that have difficulty analyzing nonlinear relationships. “In social sciences, we’re so stuck in our ways and saying, ‘My adviser ran it like this; I’m going to run it like this.’ So, we’ve been really slow to catch up.”
He says today’s students, by contrast, are entering the workforce with skills that help them to understand the mega change that AI is bringing. They are ready to “reevaluate old social science questions, making them valuable across different disciplines,” Gillion says.
Students work on a pop quiz in Daniel Gillion’s class.
Emma Luo, a fourth-year from Basking Ridge, New Jersey, majoring in communications with a minor in data science and analytics, praised the structure of the class. “There’s not a lot of busywork; it’s more focused on your understanding of the code,” she says.
The class focuses on the programming language R, which can be used easily for statistical analysis, Gillion says. “R is very accessible; they’re likely to see it in their other social science classes, so that’s another reason why I use it. We go from trying to code two plus two to turning a phrase like ‘Dr. Gillion loves math’ into an item. And then we just start slowly building.”
Each class begins with a discussion of current news. Gillion makes sure that at least one item comes from each field of sociology, economics, or political science. The goal, he says “is to take something that’s discussed at the water cooler and connect to a potential machine learning project. It’s about trying to apply it in a way that’s going to make our life better.”
Jiang says the course complements her interest in data science and how the social sciences can be deployed in the technology world. She is simultaneously taking an introductory data science course. “A lot of that course is more data visualization, doing summary statistics whereas in this course we really go into depth into all of these different models,” she says.
For the final project, students will apply a machine learning model to a topic from another class. “The skill is universal, but the application is very practical and specific to their problem,” Gillion says.
Now in its second iteration, the course has become a permanent fixture in the Department of Political Science attracting a wide range of undergraduates. “If I’m being quite honest,” Gillion says with a laugh, “if I had a course like this in my freshman year, it would have provided me with a skill that could have avoided some pain.”
Nanoparticle blueprints reveal path to smarter medicines
New research involving Penn Engineering shows detailed variation in lipid nanoparticle size, shape, and internal structure, and finds that such factors correlate with how well they deliver therapeutic cargo to a particular destination.
A generous gift from alumni Glenn and Amanda Fuhrman brings the work of internationally acclaimed artist Jaume Plensa to the University of Pennsylvania. The latest addition to the Penn Art Collection expands Philadelphia's public art.
A massive chunk of ice, a new laser, and new information on sea-level rise
For nearly a decade, Leigh Stearns and collaborators aimed a laser scanner system at Greenland’s Helheim Glacier. Their long-running survey reveals that Helheim’s massive calving events don’t behave the way scientists once thought, reframing how ice loss contributes to sea-level rise.
