Geometric Methods for Heterogeneous Data

Monday, October 7, 2013 - 10:15am - 11:05am
Keller 3-180
Susan Holmes (Stanford University)
Modern Data is currently available as combinations of heterogeneous data, medical images can be combined with genetic sequences and protein network information, the challenges of this heterogeneity involve choosing methods that can integrate multiple tables of data without loss of information, I will talk mostly about methods that use various metrics and operator approaches.
MSC Code: