KoulMde: Koul's Minimum Distance Estimation in Linear Regression and Autoregression Model by Coordinate Descent Algorithm

Consider linear regression model and autoregressive model of order q where errors in the linear regression model and innovations in the autoregression model are independent and symmetrically distributed. Hira L. Koul (1986) <doi:10.1214/aos/1176350059> proposed a nonparametric minimum distance estimation method by minimizing L2-type distance between certain weighted residual empirical processes. He also proposed a simpler version of the loss function by using symmetry of the integrating measure in the distance. Kim (2018) <doi:10.1080/00949655.2017.1392527> proposed a fast computational method which enables practitioners to compute the minimum distance estimator of the vector of general multiple regression parameters for several integrating measures. This package contains three functions: KoulLrMde(), KoulArMde(), and Koul2StageMde(). The former two provide minimum distance estimators for linear regression model and autoregression model, respectively, where both are based on Koul's method. These two functions take much less time for the computation than those based on parametric minimum distance estimation methods. Koul2StageMde() provides estimators for regression and autoregressive coefficients of linear regression model with autoregressive errors through minimum distant method of two stages. The new version is written in Rcpp and dramatically reduces computational time.

Version: 3.1.0
Depends: R (≥ 3.2.2)
Imports: Rcpp (≥ 0.12.7), expm
LinkingTo: Rcpp, RcppArmadillo
Published: 2018-05-29
Author: Jiwoong Kim
Maintainer: Jiwoong Kim <kimjiwo2 at stt.msu.edu>
License: GPL-2
NeedsCompilation: yes
CRAN checks: KoulMde results


Reference manual: KoulMde.pdf
Package source: KoulMde_3.1.0.tar.gz
Windows binaries: r-devel: KoulMde_3.1.0.zip, r-release: KoulMde_3.1.0.zip, r-oldrel: KoulMde_3.1.0.zip
OS X binaries: r-release: KoulMde_3.1.0.tgz, r-oldrel: KoulMde_3.1.0.tgz
Old sources: KoulMde archive


Please use the canonical form https://CRAN.R-project.org/package=KoulMde to link to this page.