High-dimensiona Classification for Spatially Dependent Data

Friday, February 23, 2018 - 12:20pm - 1:00pm
Lind 305
Tapabrata (Taps) Maiti (Michigan State University)
Spatial data arises in many scientific applications including image analysis and biomedical engineering. Spatial regression is an important statistical tool in dealing with spatial data interpolation and prediction. Modern applications, such as neuroimaging brings additional complexity of higher dimension and large data size. In this talk we will discuss some developments in handling spatial data with large number of feature variables and classification of a new spatial object. Our theory and numerical results indicate new developments needed.