Title: Hypergeometric-type bootstrap quasi-likelihood for functional longitudinal data: Inference and applications
Authors: Guangbao Guo - Shandong University of Technology (China) [presenting]
Abstract: Quasi-likelihood to model functional data is utilized in longitudinal settings, which is a challenging problem owing to data sparsity, time irregularity, and infinite dimensionality. We develop a novel bootstrap quasi-likelihood to derive an effective estimator. In particular, we provide a choice of sampling distribution to optimize the estimated results. Parallel bootstrap and quasi-likelihood approaches are combined to deal with irregularly and sparsely observed functional data. The proposed approach achieves some asymptotic properties under several mild conditions. The errors of the approach are smaller than those of existing approaches. Several simulation studies are conducted to illustrate the approach in the setting of discrete and finite time, using several statistical inference indicators. Finally, the excellent performance of the approach is demonstrated by analyzing real functional longitudinal data. It is shown that this approach could become popular through sampling design and parallel implementation.