Linlin Fan

Functional regression targets at the case where a scalar response is to be regressed on an infinite-dimensional functional covariate. Modeling such high-dimensional and low sample size data cast great challenge for scientists. The serious deficiency of sample size compared to the inherently infinite dimensionality causes ill-conditioning problem, which further leads to the failure in the application of traditional methods for linear model, e.g. least square estimation. Functional Data Analysis provides a whole array of innovative methods for conducting statistical analysis on functional data. Functional regression is one of the FDA methods, evaluating the dynamic association between functional covariate and response.