Detection of Influential Observations in Spatial Regression Model Based on Outliers and Bad Leverage Classification

Influential observations (IOs), which are outliers in the x direction, y direction or both, remain a problem in the classical regression model fitting. Spatial regression models have a peculiar kind of outliers because they are local in nature. Spatial regression models are also not free from the ef...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Ali Mohammed Baba, Habshah Midi, Mohd Bakri Adam, Nur Haizum Abd Rahman
Formato: article
Lenguaje:EN
Publicado: MDPI AG 2021
Materias:
Acceso en línea:https://doaj.org/article/3bcd8390d89d4cc888ff60c819805569
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:3bcd8390d89d4cc888ff60c819805569
record_format dspace
institution DOAJ
collection DOAJ
language EN
topic spatial regression model
influential observation
outlier
leverage
prediction residual
masking and swamping
Mathematics
QA1-939
spellingShingle spatial regression model
influential observation
outlier
leverage
prediction residual
masking and swamping
Mathematics
QA1-939
Ali Mohammed Baba
Habshah Midi
Mohd Bakri Adam
Nur Haizum Abd Rahman
Detection of Influential Observations in Spatial Regression Model Based on Outliers and Bad Leverage Classification
description Influential observations (IOs), which are outliers in the x direction, y direction or both, remain a problem in the classical regression model fitting. Spatial regression models have a peculiar kind of outliers because they are local in nature. Spatial regression models are also not free from the effect of influential observations. Researchers have adapted some classical regression techniques to spatial models and obtained satisfactory results. However, masking or/and swamping remains a stumbling block for such methods. In this article, we obtain a measure of spatial Studentized prediction residuals that incorporate spatial information on the dependent variable and the residuals. We propose a robust spatial diagnostic plot to classify observations into regular observations, vertical outliers, good and bad leverage points using a classification based on spatial Studentized prediction residuals and spatial diagnostic potentials, which we refer to as <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>I</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>E</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula>. Observations that fall into the vertical outliers and bad leverage points categories are referred to as IOs. Representations of some classical regression measures of diagnostic in general spatial models are presented. The commonly used diagnostic measure in spatial diagnostics, the Cook’s distance, is compared to some robust methods, <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi>H</mi><mi>i</mi><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> (using robust and non-robust measures), and our proposed <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>I</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>E</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> plots. Results of our simulation study and applications to real data showed that the Cook’s distance, non-robust <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi mathvariant="normal">H</mi><mrow><mi>si</mi><mn>1</mn></mrow><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> and robust <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi>H</mi><mrow><mi>s</mi><mi>i</mi><mn>2</mn></mrow><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> were not very successful in detecting IOs. The <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi mathvariant="normal">H</mi><mrow><mi>si</mi><mn>1</mn></mrow><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> suffered from the masking effect, and the robust <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi>H</mi><mrow><mi>s</mi><mi>i</mi><mn>2</mn></mrow><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> suffered from swamping in general spatial models. Interestingly, the results showed that the proposed <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>E</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> plot, followed by the <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>I</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> plot, was very successful in classifying observations into the correct groups, hence correctly detecting the real IOs.
format article
author Ali Mohammed Baba
Habshah Midi
Mohd Bakri Adam
Nur Haizum Abd Rahman
author_facet Ali Mohammed Baba
Habshah Midi
Mohd Bakri Adam
Nur Haizum Abd Rahman
author_sort Ali Mohammed Baba
title Detection of Influential Observations in Spatial Regression Model Based on Outliers and Bad Leverage Classification
title_short Detection of Influential Observations in Spatial Regression Model Based on Outliers and Bad Leverage Classification
title_full Detection of Influential Observations in Spatial Regression Model Based on Outliers and Bad Leverage Classification
title_fullStr Detection of Influential Observations in Spatial Regression Model Based on Outliers and Bad Leverage Classification
title_full_unstemmed Detection of Influential Observations in Spatial Regression Model Based on Outliers and Bad Leverage Classification
title_sort detection of influential observations in spatial regression model based on outliers and bad leverage classification
publisher MDPI AG
publishDate 2021
url https://doaj.org/article/3bcd8390d89d4cc888ff60c819805569
work_keys_str_mv AT alimohammedbaba detectionofinfluentialobservationsinspatialregressionmodelbasedonoutliersandbadleverageclassification
AT habshahmidi detectionofinfluentialobservationsinspatialregressionmodelbasedonoutliersandbadleverageclassification
AT mohdbakriadam detectionofinfluentialobservationsinspatialregressionmodelbasedonoutliersandbadleverageclassification
AT nurhaizumabdrahman detectionofinfluentialobservationsinspatialregressionmodelbasedonoutliersandbadleverageclassification
_version_ 1718410271419006976
spelling oai:doaj.org-article:3bcd8390d89d4cc888ff60c8198055692021-11-25T19:06:10ZDetection of Influential Observations in Spatial Regression Model Based on Outliers and Bad Leverage Classification10.3390/sym131120302073-8994https://doaj.org/article/3bcd8390d89d4cc888ff60c8198055692021-10-01T00:00:00Zhttps://www.mdpi.com/2073-8994/13/11/2030https://doaj.org/toc/2073-8994Influential observations (IOs), which are outliers in the x direction, y direction or both, remain a problem in the classical regression model fitting. Spatial regression models have a peculiar kind of outliers because they are local in nature. Spatial regression models are also not free from the effect of influential observations. Researchers have adapted some classical regression techniques to spatial models and obtained satisfactory results. However, masking or/and swamping remains a stumbling block for such methods. In this article, we obtain a measure of spatial Studentized prediction residuals that incorporate spatial information on the dependent variable and the residuals. We propose a robust spatial diagnostic plot to classify observations into regular observations, vertical outliers, good and bad leverage points using a classification based on spatial Studentized prediction residuals and spatial diagnostic potentials, which we refer to as <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>I</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>E</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula>. Observations that fall into the vertical outliers and bad leverage points categories are referred to as IOs. Representations of some classical regression measures of diagnostic in general spatial models are presented. The commonly used diagnostic measure in spatial diagnostics, the Cook’s distance, is compared to some robust methods, <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi>H</mi><mi>i</mi><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> (using robust and non-robust measures), and our proposed <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>I</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>E</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> plots. Results of our simulation study and applications to real data showed that the Cook’s distance, non-robust <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi mathvariant="normal">H</mi><mrow><mi>si</mi><mn>1</mn></mrow><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> and robust <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi>H</mi><mrow><mi>s</mi><mi>i</mi><mn>2</mn></mrow><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> were not very successful in detecting IOs. The <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi mathvariant="normal">H</mi><mrow><mi>si</mi><mn>1</mn></mrow><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> suffered from the masking effect, and the robust <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msubsup><mi>H</mi><mrow><mi>s</mi><mi>i</mi><mn>2</mn></mrow><mn>2</mn></msubsup></mrow></semantics></math></inline-formula> suffered from swamping in general spatial models. Interestingly, the results showed that the proposed <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>E</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> plot, followed by the <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>I</mi><mi>S</mi><mi>R</mi><mi>s</mi><mo>−</mo><msub><mi>P</mi><mrow><mi>o</mi><mi>s</mi><mi>i</mi></mrow></msub></mrow></semantics></math></inline-formula> plot, was very successful in classifying observations into the correct groups, hence correctly detecting the real IOs.Ali Mohammed BabaHabshah MidiMohd Bakri AdamNur Haizum Abd RahmanMDPI AGarticlespatial regression modelinfluential observationoutlierleverageprediction residualmasking and swampingMathematicsQA1-939ENSymmetry, Vol 13, Iss 2030, p 2030 (2021)