He’s going to predict your next favorite song

Hunter Osterhoudt wants to predict your next favorite song. A master’s student in the School of Computing and Information, Osterhoudt (SCI ’20) is pursuing a graduate degree with an emphasis on machine learning and data science. He hopes it’s the gateway to a career at a streaming platform and an opportunity to meld his two passions — music and computing.

While some might not make an intuitive connection between technology and the arts, it’s a natural pairing for Osterhoudt: his introduction to playing music was with a guitar-shaped video game controller.

As a child, he was so enamored with the game Guitar Hero 3 that his parents bought him a real guitar. Today, Osterhoudt is a singer and multi-instrumentalist playing electric bass, ukulele, mandolin and keyboard in addition to his first love.

Forging a path

When it came to choosing a graduate program, Osterhoudt saw the advantage of remaining at Pitt to pursue an advanced degree.

“As I was wrapping up undergrad, I felt conflicted about what I wanted to do next. Though I felt ready to get a job and start my professional life, I still felt there was more that I wanted to learn and experience in my field,” he said.

At his advisor’s urging, he took a look at the Pitt2Pitt Scholarship as a means to pay for graduate school. The program, launched in 2020, provides financial incentives — up to $7,500 to one of 84 Pitt master’s programs — and support to all Pitt undergraduate students and alumni seeking to continue their education at their alma mater.

For students undertaking a graduate degree in the School of Computing and Information (SCI), their course of study is two-fold, said Emily Bennett, manager of experiential learning in SCI.

“They are not only building a foundational knowledge of their field of study, but they are also enterprising next-level research with real-world applications,” she said.

In a machine-learning class, for example, Osterhoudt first researched the algorithms and the audio signal processing that power music genre classification.

“In undergrad, you’re typically assigned your projects, but in graduate school, you gain the liberty and competence to explore your interests and execute your ideas,” said Osterhoudt. “My class equipped me with the tools to be able to branch out and investigate the various technology companies use in their recommendation systems.”

Though the specific algorithms used by companies like Apple Music and Pandora are held close as corporate intellectual property, Osterhoudt was able to create his own program.

And machine learning isn’t the only class where Osterhoudt acquired skills applicable to a career at a streaming platform.

His graduate classes also delved into text and computer architecture, spectrograms and natural language processing — additional elements that make up the science of AI.

“An advantage of taking classes in the School of Computing and Information is the diversity in professors’ expertise and teaching style,” said Osterhoudt. “Each course is an opportunity to approach learning from a different angle.”

What’s next

After completing his final semester of coursework, Osterhoudt is embarking on the last step to earning his degree. Computer science graduate students can complete a master’s thesis or a master’s project.

Some students choose the six-credit thesis option so they can delve more deeply into a particular research topic, whereas others choose to complete a three-credit project, making room in their schedule to take an additional class to add to the breadth of their academic undertaking.

Though he hasn’t yet chosen the focus of his project, Osterhoudt is certain he’ll be taking a hands-on approach to learning.

“My graduate school experience has taught me how to approach challenges and apply my knowledge. Now I can identify a problem [in computer science] and articulate the steps to addressing it.”

— Nichole Faina, photography by Tom Altany