When industrial engineering students hit the job market, employers are after candidates who can comfortably manage, analyze and act on data.
The Department of Industrial and Systems Engineering at the University of Wisconsin-Madison has heard that feedback—from alumni and employers alike—loud and clear. And it’s responded by creating an array of courses in recent years to impart those in-demand skills to both undergraduate and graduate students.
ISyE’s set of data science courses introduce students to modern tools and techniques, while also presenting the type of systems-level viewpoint that’s a hallmark of an industrial engineering education. In the past few years, ISyE has created four courses that blend technical skill with the broad perspective that’s essential to responsibly using data:
- ISyE 210: Introduction to Industrial Statistics
- ISyE 521: Machine Learning in Action for Industrial Engineers
- ISyE 562: Human Factors of Data Science and Machine Learning
- ISyE 649: Interactive Data Analytics
The department has also partnered with the Department of Electrical and Computer Engineering to develop a fifth course, ISyE/ECE 570: Ethics of Data for Engineers.
All five are open to students from across the college, regardless of engineering major (with completion of prerequisites).
A common theme runs across them: Less time on theory, more on learning by actively applying techniques to practical problems.
“If you have seen this material in action, you know that it works,” says Amanda Smith, an assistant teaching professor and associate chair for undergraduate affairs who developed Introduction to Industrial Statistics. “I think students will care more about learning it, because they will see that it is relevant to what they are going to be doing.”
Assistant Professor Justin Boutilier, who developed and teaches Machine Learning in Action for Industrial Engineers, starts each lecture with a relevant use case to emphasize a given method’s utility, such as predicting U.S. Supreme Court decisions and teasing out the risk factors associated with heart disease.
“I hope that it just shows them how these methods can actually be used in interesting problems, and that they know enough to go do this,” says Boutilier.
Students in Interactive Data Analytics, an evolution of a course developed by Emerson Electric Quality & Productivity Professor John Lee, learn the programming language R and create a data visualization around a topic of their choosing.
“Data visualization is one of the major ways in which you translate data into information that can be acted upon,” says Assistant Professor Tony McDonald, who’s teaching the course in fall 2023. “Data visualization tends to be a step that either gets skipped or has minimal attention put on it, because of the way that machine learning algorithm development has become so automated. In many cases, those skipped data visualization steps are ones that could have prevented poor deployment outcomes for machine learning-based technology.”
That spirit of viewing data with intention and consideration of both its flaws and potential ramifications is at the core of Ethics of Data for Engineers, a new course that is the capstone of the college’s new undergraduate certificate in engineering data analytics (a joint offering between ISyE and ECE). Kangwook Lee, an assistant professor of electrical and computer engineering, is developing and will teach the course in spring 2024.
Those ideas are also central to John Lee’s Human Factors of Data Science and Machine Learning, which delves into questions of representative sampling, bias, fairness, trust and more through case studies featuring leading tech companies.
“These algorithms should be designed with sensitivity to human values and the consequence for their application in the world. Which I think is so easy to lose sight of when you’re creating something and you’re just happy when it runs,” says Lee. “It’s so easy to forget the 10-20% of the cases where it doesn’t work well. Then also thinking about, why are we doing this? What is the purpose of this? Going beyond, can we do it technically? But thinking about, why are we going down this path? What is the benefit to people and society?”
Top image: Undergraduate and graduate students critique data visualizations in ISyE 649: Interactive Data Analytics, giving them a chance to learn how to effectively present data. Photo by: Tom Ziemer.