Mustafa Bilgic, assistant professor of computer science, has received a five-year, $549,863 National Science Foundation (NSF) CAREER Award for “Active Learning through Rich and Transparent Interactions.”
The prestigious CAREER Awards support early career development activities of junior faculty who exemplify the role of teacher-scholars through the integration of outstanding research and excellent education. Associate Professors Cindy Hood and Shangping Ren and Assistant Professor Ioan Raicu previously won the award in the Computer Science Department.
Many day-to-day tasks are now powered by machine learning algorithms. Examples include web search, speech recognition, product recommendations, fraud detection, and medical diagnosis to name a few. Building machine learning models often requires human supervision (e.g., rating movies and providing example fraud cases), which costs time and effort.
Bilgic and his team are working on developing active learning algorithms that are aimed at minimizing human supervision costs while maximizing accuracy of the algorithms. Under the CAREER project, Bilgic proposes a new active learning paradigm in which the algorithms will be enabled to exchange a rich set of questions and answers with users. For example, for building a fraud detection model, algorithms and experts will work together and discuss various cases in detail, exchanging questions and answers. This “rich and transparent” paradigm will lower the costs associated with human supervision, increase users’ trust in the trained models, and enable faster development of intelligent systems.
Two Ph.D. and several undergraduate and high school students will be trained under this award. A new graduate course on interactive machine learning will be developed. Bilgic also has partnered with a Chicago public high school for outreach.
Bilgic directs the Machine Learning Laboratory at IIT.