A model where the least trimmed squares estimator is maximum likelihood
<p>The least trimmed squares (LTS) estimator is a popular robust regression estimator. It finds a subsample of <em>h</em> ‘good’ observations among <em>n</em> observations and applies least squares on that subsampl...
Main Authors: | , , |
---|---|
Format: | Journal article |
Language: | English |
Published: |
Wiley
2023
|
_version_ | 1811139239487733760 |
---|---|
author | Berenguer-Rico, V Johansen, S Nielsen, B |
author_facet | Berenguer-Rico, V Johansen, S Nielsen, B |
author_sort | Berenguer-Rico, V |
collection | OXFORD |
description | <p>The least trimmed squares (LTS) estimator is a popular robust regression estimator. It finds a subsample of <em>h</em> ‘good’ observations among <em>n</em> observations and applies least squares on that subsample. We formulate a model in which this estimator is maximum likelihood. The model has ‘outliers’ of a new type, where the outlying observations are drawn from a distribution with values outside the realized range of <em>h</em> ‘good’, normal observations. The LTS estimator is found to be <em>h</em><sup>1/2</sup> consistent and asymptotically standard normal in the location-scale case. Consistent estimation of <em>h</em> is discussed. The model differs from the commonly used <em>ϵ</em>-contamination models and opens the door for statistical discussion on contamination schemes, new methodological developments on tests for contamination as well as inferences based on the estimated good data.</p> |
first_indexed | 2024-03-07T07:53:52Z |
format | Journal article |
id | oxford-uuid:039f0f75-72ac-4059-80e5-17d03fb19d25 |
institution | University of Oxford |
language | English |
last_indexed | 2024-09-25T04:02:56Z |
publishDate | 2023 |
publisher | Wiley |
record_format | dspace |
spelling | oxford-uuid:039f0f75-72ac-4059-80e5-17d03fb19d252024-05-15T10:12:01ZA model where the least trimmed squares estimator is maximum likelihoodJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:039f0f75-72ac-4059-80e5-17d03fb19d25EnglishSymplectic ElementsWiley2023Berenguer-Rico, VJohansen, SNielsen, B<p>The least trimmed squares (LTS) estimator is a popular robust regression estimator. It finds a subsample of <em>h</em> ‘good’ observations among <em>n</em> observations and applies least squares on that subsample. We formulate a model in which this estimator is maximum likelihood. The model has ‘outliers’ of a new type, where the outlying observations are drawn from a distribution with values outside the realized range of <em>h</em> ‘good’, normal observations. The LTS estimator is found to be <em>h</em><sup>1/2</sup> consistent and asymptotically standard normal in the location-scale case. Consistent estimation of <em>h</em> is discussed. The model differs from the commonly used <em>ϵ</em>-contamination models and opens the door for statistical discussion on contamination schemes, new methodological developments on tests for contamination as well as inferences based on the estimated good data.</p> |
spellingShingle | Berenguer-Rico, V Johansen, S Nielsen, B A model where the least trimmed squares estimator is maximum likelihood |
title | A model where the least trimmed squares estimator is maximum likelihood |
title_full | A model where the least trimmed squares estimator is maximum likelihood |
title_fullStr | A model where the least trimmed squares estimator is maximum likelihood |
title_full_unstemmed | A model where the least trimmed squares estimator is maximum likelihood |
title_short | A model where the least trimmed squares estimator is maximum likelihood |
title_sort | model where the least trimmed squares estimator is maximum likelihood |
work_keys_str_mv | AT berenguerricov amodelwheretheleasttrimmedsquaresestimatorismaximumlikelihood AT johansens amodelwheretheleasttrimmedsquaresestimatorismaximumlikelihood AT nielsenb amodelwheretheleasttrimmedsquaresestimatorismaximumlikelihood AT berenguerricov modelwheretheleasttrimmedsquaresestimatorismaximumlikelihood AT johansens modelwheretheleasttrimmedsquaresestimatorismaximumlikelihood AT nielsenb modelwheretheleasttrimmedsquaresestimatorismaximumlikelihood |