Pang Du, Ph.D., professor in the Department of Statistics at Virginia Tech, will give a talk in the Department of Mathematical Sciences Statistics Seminar on Friday, Oct. 16, 2026, from 11:30 a.m. to 12:30 p.m. in HOS 380. The title of his talk is “Regression Modeling of Zero-Inflated Longitudinal Functional Data.” Refreshments will be served at 11:15 a.m.

Abstract: Zero-inflated functional data appear when an excessive number of zeros are recorded in functional measurements, either because the true value is zero or because values fall below the limit of detection. The longitudinal version of such data has the extra complication of repeated measures per subject. We propose a two-part mixed-effects functional regression model for such zero-inflated longitudinal functional data. The first part models the probability that the functional response is nonzero using a mixed-effects functional logistic regression model. The second part models the log of response function, conditional on being positive, using a mixed-effects functional linear model. Smooth fixed effects are represented by spline bases and estimated with roughness penalties. Parameters are estimated via penalized quasi-likelihood for the binary part and a REML-based EM algorithm for the nonzero part. Extensive simulations are presented to evaluate the numerical performance of our method. We also apply the method to a Northwestern ICU study to investigate the relationship between total calcium and albumin measurements in repeated blood tests during each of the multiple ICU visits per patient. Results show that the proposed approach effectively handles zero inflation while recovering the functional relationship between the variables of interest. This is joint work with Anbin Rhee and Wei Zhang.