Using proposed nonparametric regression models for clustered data : a simulation study

Abstract EN

A nonparametric model is introduced to account varying impacts of factors over clusters using proposed models in comparison with some classical models.

It achieves the parsimony of parameterization and allows the explorations of nonlinear interactions.

The random effect in the nonparametric model also accounts within-cluster correlation.

Local, linear-based estimation procedure is proposed for estimating functional coefficients, residual variance and within-cluster correlation matrix for three cases (the autoregressive, the exchangeable and the unstructured).

Simulation studies are carried out to demonstrate the methodological power of the proposed methods in the finite sample, Using MATLAB language program designed for this purpose