I am a Postdoc Researcher in the Machine Learning and Perception group at Microsoft Research Cambridge. My research focuses on online learning. I am convinced that future interactive systems will have to learn directly from their users, and adapt to their preferences. I develop exploration, reward, and learning approaches for making this possible.
Before joining Microsoft Research, I did my PhD in the ILPS group at the University of Amsterdam. My research focused on developing efficient and robust online learning to rank methods, that can learn effectively from noisy, relative user feedback. These methods are applicable to developing “self-learning search engines” that can automatically learn by interacting with their users. My advisors were Maarten de Rijke and Shimon Whiteson.
Previously
I received my MSc in Computer Science from California State University, East Bay. At CSU East Bay I was member of the research lab Adaptive Hypermedia and Assistive Technologies lab (AHAT). I completed my thesis research was on Subsymbolic User Modeling in Adaptive Hypermedia under the guidance of Dr. Hilary Holz. Before that, I obtained the degree of Diplom-Informatiker (FH) from University of Applied Sciences (Hochschule für Technik und Wirtschaft), Dresden, Germany in 2003.
Katja Hofmann 