Bioinformatics Seminar

 

Wednesday, April 28, 2010

3:00 - 4:00

Room 457 Blocker

 

Vadim Zipunnikov

Department of Biostatistics

Johns Hopkins Bloomberg

School of Public Health

 

 

 

Multilevel Functional Principal Component Analysis for High-Dimensional Data (HD-MFPCA).

 

Massive and complex data sets raise a host of statistical challenges that were unthinkable even a few years ago. An example is the statistical analysis of large samples of high dimensional vectors observed at multiple visits when even loading a few vectors in the computer memory is impossible. Our approach provides a fast solution that circumvents the many computational and inferential problems associated with high dimensional data in this context. Our methods are applied to a large study of imaging where MRI data are available for hundreds of subjects at multiple visits. Our methods are applicable to any type of study where data can be unfolded into a long vector including densely observed functions and images.