Jim Thompson is a big deal in the tech world; the chief technology officer of Qualcomm Technologies—the multinational mobile chip, semiconductor and software company—is the type of speaker who delivers keynotes at IEEE conferences and advises the National Science Foundation on engineering issues.
But in July 2021, Thompson (BSEE ‘85, MSEE ‘87, PhD EE ‘91), took time to digitally drop by a session of ECE 697, the capstone course of the accelerated master’s degree in machine learning and signal processing. For an hour, Thompson discussed his journey through academia, his time in the technology sector, and how he managed to stay on the cutting edge through decades of rapid change.
The interaction was one of many career-building experiences students in the program received. Unlike a traditional master’s degree, which focuses on academic research and culminates in a research-based thesis, the accelerated master’s degree program is a course-based curriculum that takes 12 to 16 months to complete. The program is designed for students hoping to get a jump start on a career in data science or who are returning to school to improve their skills.
Thompson was one of several speakers that visited with the class, including engineers from GE Healthcare, American Family Insurance, Google and Amazon. “The goal is to get a broad perspective from engineers that are at different points in their career,” says Matt Malloy, an assistant teaching professor in electrical and computer engineering, who led the class. “The students also learn about various industries where machine learning and signal processing are fundamental. And the perspective from both practicing engineers and executives is invaluable when you’re figuring out what option to pursue after you graduate.”
Most students that join the program, Malloy says, know that they want to pursue opportunities in industry or have a few years of experience, but don’t want to lose two full years of income with a research-based master’s program. The accelerated approach is a good alternative, he says, and within ECE, it is among other specialized degrees, including an accelerated master’s degree for professionals and an online power engineering degree.
“For students just finishing with a BS, the draw is that they’re at a point where they have just finished learning the fundamentals and they can now take all these cool upper-level courses that are really more focused and more applied,” Malloy says. “If you don’t stick around and get a master’s, you kind of leave all that all on the table.”
It also means missing out on some great hands-on experiences. In ECE 697, the capstone design course in machine learning and signal processing, students pursue final projects that tie together their previous year of coursework. In summer 2021, students pursued projects with real-life applications, including a computer vision system that reads Chinese license plates, a convolutional neural network that monitors the quality of spot welds, a short-term stock market predictor, an artificial neural network that can translate MRI scans into CT scans and a phrase detector keyed to the words, “Hey Bucky!”
That learning-by-doing approach helps the students drill down to exactly where they want to take their career—and most importantly, shows them how to be adaptable. That’s something Thompson emphasized in his talk. “What happens in your career is that you’re going to get a task you end up focusing on, let’s say video processing. And you’re going to become an expert in a couple of years,” he said. “But then you might do something else. And it’s those fundamental, base kind of skills that will allow you to move between different types of signal processing that are really important.”
Author: Jason Daley