Seminar: In-situ Computing
Speaker: Michael Coen, Department of Biostatistics & Medical Informatics and Department of Computer Science, University of Wisconsin-Madison
Date: October 17, 2011
Time: 4:10 p.m.
Location: 1312 Hoover Hall
Abstract: My work addresses a new framework for data analysis that bridges statistics, optimization theory, and machine learning. I introduce a family of computational methods that are heavily tied to the shapes and densities of datasets. This approach is visual and intuitive, while being mathematically formal, non-parametric, and distribution free. Applications are broad, including: medical informatics and imaging; clustering; protein homology; spatial-based kernels; and a new class of highly powerful goodness-of-fit and hypothesis tests. This talk will focus on both the theory and applications of this work, specifically examining some of its novel mathematical techniques. No prior background is assumed. (This work is joint with Hidayath Ansari, Marissa Phillips, and Timothy Chang.)
Speaker bio: Mike Coen is an assistant professor at UW-Madison. His primary research interest is self-supervised machine learning, based on biologically inspired models of learning in animals. His Ph.D., S.M, and S.B. degrees are from MIT, where his doctoral thesis received the Sprowls Award for best dissertation in Computer Science.
This lecture was made possible in part by the generosity of F. Wendell Miller, who left his entire estate jointly to Iowa State University and the University of Iowa. Mr. Miller, who died in 1995 at age 97, was born in Altoona, Illinois, grew up in Rockwell City, graduated from Grinnell College and Harvard Law School and practiced law in Des Moines and Chicago before returning to Rockwell City to manage his family’s farm holdings and to practice law. His will helped to establish the F. Wendell Miller Trust, the annual earnings on which, in part, helped to support this activity.