Empirical Likelihood Inference for a Partially Linear Model under Longitudinal Data

Yu Ying Jiang

doi:10.4028/www.scientific.net/AMM.353-356.3355

Paper Titles

The Internal Stress Optimization of Structural Model Based on Super Finite Element Calculating Software
p.3334

Numerical Study on Airfoil Diversion Technology Used in Saving Energy by Air Injection on Large Ship
p.3341

Numerical Analysis of Viscous-Spring Artificial Boundary Element for Underground Explosion
p.3345

Another Solving Equation of Elastoplastic Problem Based on Stress Strain Relation
p.3350

Empirical Likelihood Inference for a Partially Linear Model under Longitudinal Data
p.3355

An Improved Dynamic Programming Method for Solving the Problem of Nonlinear Programming
p.3359

Energy Flow Finite Element Analysis of L-Shaped Plate Including Three Types of Waves
p.3365

A Coupled SBFEM-FEM Approach for Evaluating Stress Intensity Factors
p.3369

The Finite Element Model Updating Based on Multi-Case Displacement Measurement
p.3378

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vols. 353-356Empirical Likelihood Inference for a Partially...

Empirical Likelihood Inference for a Partially Linear Model under Longitudinal Data

Abstract:

In this paper, a partially linear model under longitudinal data is considered. In order to take into consideration the within-subject correlation structure of the repeated measurements, an empirical likelihood incorporating the correlation structure is developed. The asymptotic normality of the maximum empirical likelihood estimates of the regression coefficients is obtained. It also can be shown that the proposed empirical likelihood ratio is asymptotically standard chi-square. The results can be used directly to construct the asymptotic confidence regions of the regression coefficients. The convergence rate of the baseline function is derived.