Activity Number:
|
257
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 13, 2002 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistical Computing*
|
Abstract - #300943 |
Title:
|
Reparametrization in Markov Chain Marginal Bootstrap
|
Author(s):
|
Maria Kocherginsky*+ and Xuming He
|
Affiliation(s):
|
University of Illinois, Urbana-Champaign and University of Illinois, Urbana-Champaign
|
Address:
|
101 Illini Hall, 725 S Wright St., Champaign, Illinois, 61820, USA
|
Keywords:
|
Bootstrap ; Markov Chain ; M-estimator ; Nonlinear regression ; Transformation
|
Abstract:
|
Markov chain marginal bootstrap (MCMB) is a new bootstrap method proposed by He and Hu (2002) for constructing confidence intervals or regions based on likelihood equations. It is designed to ease the computational burden of bootstrap in high-dimensional problems. It differs from the usual bootstrap methods in two aspects: a set of p one-dimensional equations is solved in place of solving a p-dimensional system of equations for each bootstrap estimate of the parameter; the resulting estimates form a Markov chain rather than an independent sequence of realizations. This paper discusses the use of MCMB for nonlinear regression models and proposes an appropriate reparameterization, which eliminates high autocorrelation of the resulting chains and improves on the efficiency and stability of the procedure.
|