Title: Fast multilevel functional principal component analysis
Authors: Erjia Cui - Johns Hopkins University (United States) [presenting]
Luo Xiao - North Carolina State University (United States)
Ciprian Crainiceanu - Johns Hopkins University (United States)
Ruonan Li - North Carolina State University (United States)
Abstract: Fast multilevel functional principal component analysis (fast MFPCA) is introduced, which scales up to high dimensional functional data measured at multiple visits. The new approach is orders of magnitude faster than and achieves comparable estimation accuracy with the original MFPCA. Methods are motivated by the National Health and Nutritional Examination Survey (NHANES), which contains minute-level physical activity information of more than 10000 participants over multiple days and 1440 observations per day. While MFPCA takes more than five days to analyze these data, fast MFPCA takes less than five minutes. A theoretical study of the proposed method is also provided. The associated function mfpca.face() is available in the R package refund.