Recent Advances in the Analysis of Massive Spatial Data
Hao Zhang, Department of Statistics, Purdue University
Monday, February 1, 2010
4:00 p.m., 223 Weber
The analysis of spatial data collected at a large number of locations and possibly from different sources presents great challenges to statistical inferences. One of the challenges is in dealing with the large covariance matrix and its inverse. Two dominating approaches to overcoming this large matrix issue are to exploit the computational advantage of sparse matrices and to impose some low rank structure. I will present some recent theoretical and computational developments in both approaches.