Robust mixture regression model fitting by Laplace distribution

Master of Science === Department of Statistics === Weixing Song === A robust estimation procedure for mixture linear regression models is proposed in this report by assuming the error terms follow a Laplace distribution. EM algorithm is imple- mented to conduct the estimation procedure of missing...

Full description

Bibliographic Details
Main Author:	Xing, Yanru
Language:	en_US
Published:	Kansas State University 2013
Subjects:	EM algorithm Laplace distribution Least absolute deviation Mixture regression model Statistics (0463)
Online Access:	http://hdl.handle.net/2097/16534

id	ndltd-KSU-oai-krex.k-state.edu-2097-16534
record_format	oai_dc
spelling	ndltd-KSU-oai-krex.k-state.edu-2097-165342017-03-03T15:45:08Z Robust mixture regression model fitting by Laplace distribution Xing, Yanru EM algorithm Laplace distribution Least absolute deviation Mixture regression model Statistics (0463) Master of Science Department of Statistics Weixing Song A robust estimation procedure for mixture linear regression models is proposed in this report by assuming the error terms follow a Laplace distribution. EM algorithm is imple- mented to conduct the estimation procedure of missing information based on the fact that the Laplace distribution is a scale mixture of normal and a latent distribution. Finite sample performance of the proposed algorithm is evaluated by some extensive simulation studies, together with the comparisons made with other existing procedures in this literature. A sensitivity study is also conducted based on a real data example to illustrate the application of the proposed method. 2013-09-27T19:06:21Z 2013-09-27T19:06:21Z 2013-09-27 2013 December Report http://hdl.handle.net/2097/16534 en_US Kansas State University
collection	NDLTD
language	en_US
sources	NDLTD
topic	EM algorithm Laplace distribution Least absolute deviation Mixture regression model Statistics (0463)
spellingShingle	EM algorithm Laplace distribution Least absolute deviation Mixture regression model Statistics (0463) Xing, Yanru Robust mixture regression model fitting by Laplace distribution
description	Master of Science === Department of Statistics === Weixing Song === A robust estimation procedure for mixture linear regression models is proposed in this report by assuming the error terms follow a Laplace distribution. EM algorithm is imple- mented to conduct the estimation procedure of missing information based on the fact that the Laplace distribution is a scale mixture of normal and a latent distribution. Finite sample performance of the proposed algorithm is evaluated by some extensive simulation studies, together with the comparisons made with other existing procedures in this literature. A sensitivity study is also conducted based on a real data example to illustrate the application of the proposed method.
author	Xing, Yanru
author_facet	Xing, Yanru
author_sort	Xing, Yanru
title	Robust mixture regression model fitting by Laplace distribution
title_short	Robust mixture regression model fitting by Laplace distribution
title_full	Robust mixture regression model fitting by Laplace distribution
title_fullStr	Robust mixture regression model fitting by Laplace distribution
title_full_unstemmed	Robust mixture regression model fitting by Laplace distribution
title_sort	robust mixture regression model fitting by laplace distribution
publisher	Kansas State University
publishDate	2013
url	http://hdl.handle.net/2097/16534
work_keys_str_mv	AT xingyanru robustmixtureregressionmodelfittingbylaplacedistribution
_version_	1718418551605297152

Robust mixture regression model fitting by Laplace distribution

Similar Items