×

A nonparametric regression model for panel count data analysis. (English) Zbl 1429.62152

The authors consider panel count data, occurring at some discrete time points, which are not exactly known. In order to take into account nonlinear covariate effects arising due to certain non-controllable variables, nonparametric regression models are considered for panel count data. The regression B-spline method is applied to model and estimate the regression function of covariates and the reference (baseline) mean function. Asymptotic properties of the resulting estimation procedure are considered under certain distributional assumptions, and simulation studies are provided. Moreover, applications are given to a childhood wheezing study.

MSC:

62G08 Nonparametric regression and quantile regression
65D07 Numerical computation using splines
62D20 Causal inference from observational studies
62P10 Applications of statistics to biology and medical sciences; meta analysis