Colin Raffel has officially joined the faculty of the Department of Computer Science as an assistant professor. Raffel received a doctorate in electrical engineering from Columbia University in 2016 and spent the last four years working with Google Brain.
Raffel’s research focuses on developing machine learning techniques, especially semi-supervised, unsupervised, and transfer learning methods for learning from limited labeled data. His work addresses the shortcomings of machine learning that hold back its adoption in practically useful settings, and much of his recent research has been concerned with making machine learning methods less reliant on large labeled datasets to attain good performance. There are many applications where these methods could be useful, but obtaining labeled data is prohibitively time- or resource-intensive. Colin hopes to find solutions that bring machine learning into as many new applications as possible, and he welcomes potential research challenges from anyone who has been unable to apply it to an important problem because of some major shortcoming.
Now that he is in Chapel Hill, Raffel is excited to build a lab and mentor students, something he had limited opportunities to do during his four years in industry.
“I’ve always enjoyed the back-and-forth involved with advising students, where in the best case the advisor learns as much as the student,” Raffel said. “The students I met during my visits were extremely sharp and engaged.”
Raffel said he was drawn to UNC CS because of the department culture and the support system for new faculty. Operations this semester will be far from normal, but Raffel is confident that this is the place to be while he builds a research group and tackles new challenges.