A weighted least squares procedure to approximate least absolute deviation estimation in time series...
A weighted least squares procedure to approximate least absolute deviation estimation in time series with specific reference to infinite variance unit root problems
About this item
Full title
Author / Creator
Publisher
Ithaca: Cornell University Library, arXiv.org
Journal title
Language
English
Formats
Publication information
Publisher
Ithaca: Cornell University Library, arXiv.org
Subjects
More information
Scope and Contents
Contents
A weighted regression procedure is proposed for regression type problems where the innovations are heavy-tailed. This method approximates the least absolute regression method in large samples, and the main advantage will be if the sample is large and for problems with many independent variables. In such problems bootstrap methods must often be util...
Alternative Titles
Full title
A weighted least squares procedure to approximate least absolute deviation estimation in time series with specific reference to infinite variance unit root problems
Authors, Artists and Contributors
Author / Creator
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_proquest_journals_2084751053
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_proquest_journals_2084751053
Other Identifiers
E-ISSN
2331-8422
