Cellwise Outlier Detection in Heterogeneous Populations

<p>Real-world applications may be affected by outlying values. In the model-based clustering literature, several methodologies have been proposed to detect units that deviate from the majority of the data (rowwise outliers) and trim them from the parameter estimates. However, the discarded obs...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Giorgia Zaccaria (21245577) (author)
مؤلفون آخرون: Luis A. García-Escudero (21245580) (author), Francesca Greselin (14203066) (author), Agustín Mayo-Íscar (9571212) (author)
منشور في: 2025
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1852018897395384320
author Giorgia Zaccaria (21245577)
author2 Luis A. García-Escudero (21245580)
Francesca Greselin (14203066)
Agustín Mayo-Íscar (9571212)
author2_role author
author
author
author_facet Giorgia Zaccaria (21245577)
Luis A. García-Escudero (21245580)
Francesca Greselin (14203066)
Agustín Mayo-Íscar (9571212)
author_role author
dc.creator.none.fl_str_mv Giorgia Zaccaria (21245577)
Luis A. García-Escudero (21245580)
Francesca Greselin (14203066)
Agustín Mayo-Íscar (9571212)
dc.date.none.fl_str_mv 2025-06-30T13:40:07Z
dc.identifier.none.fl_str_mv 10.6084/m9.figshare.28931076.v2
dc.relation.none.fl_str_mv https://figshare.com/articles/dataset/Cellwise_outlier_detection_in_heterogeneous_populations/28931076
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Biotechnology
Ecology
Cancer
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
Cellwise contamination
EM algorithm
Imputation
Missing data
Model-based clustering
Robustness
dc.title.none.fl_str_mv Cellwise Outlier Detection in Heterogeneous Populations
dc.type.none.fl_str_mv Dataset
info:eu-repo/semantics/publishedVersion
dataset
description <p>Real-world applications may be affected by outlying values. In the model-based clustering literature, several methodologies have been proposed to detect units that deviate from the majority of the data (rowwise outliers) and trim them from the parameter estimates. However, the discarded observations can encompass valuable information in some observed features. Following the more recent cellwise contamination paradigm, we introduce a Gaussian mixture model for cellwise outlier detection. The proposal is estimated via an Expectation-Maximization (EM) algorithm with an additional step for flagging the contaminated <i>cells</i> of a data matrix and then imputing—instead of discarding—them before the parameter estimation. This procedure adheres to the spirit of the EM algorithm by treating the contaminated cells as missing values. We analyze the performance of the proposed model in comparison with other existing methodologies through a simulation study with different scenarios and illustrate its potential use for clustering, outlier detection, and imputation on three real datasets. Additional applications include socio-economic studies, environmental analysis, healthcare, and any domain where the aim is to cluster data affected by missing information and outlying values within features.</p>
eu_rights_str_mv openAccess
id Manara_fcc03421f03fe792f5be2a973b8580c8
identifier_str_mv 10.6084/m9.figshare.28931076.v2
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/28931076
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Cellwise Outlier Detection in Heterogeneous PopulationsGiorgia Zaccaria (21245577)Luis A. García-Escudero (21245580)Francesca Greselin (14203066)Agustín Mayo-Íscar (9571212)BiotechnologyEcologyCancerMathematical Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedCellwise contaminationEM algorithmImputationMissing dataModel-based clusteringRobustness<p>Real-world applications may be affected by outlying values. In the model-based clustering literature, several methodologies have been proposed to detect units that deviate from the majority of the data (rowwise outliers) and trim them from the parameter estimates. However, the discarded observations can encompass valuable information in some observed features. Following the more recent cellwise contamination paradigm, we introduce a Gaussian mixture model for cellwise outlier detection. The proposal is estimated via an Expectation-Maximization (EM) algorithm with an additional step for flagging the contaminated <i>cells</i> of a data matrix and then imputing—instead of discarding—them before the parameter estimation. This procedure adheres to the spirit of the EM algorithm by treating the contaminated cells as missing values. We analyze the performance of the proposed model in comparison with other existing methodologies through a simulation study with different scenarios and illustrate its potential use for clustering, outlier detection, and imputation on three real datasets. Additional applications include socio-economic studies, environmental analysis, healthcare, and any domain where the aim is to cluster data affected by missing information and outlying values within features.</p>2025-06-30T13:40:07ZDatasetinfo:eu-repo/semantics/publishedVersiondataset10.6084/m9.figshare.28931076.v2https://figshare.com/articles/dataset/Cellwise_outlier_detection_in_heterogeneous_populations/28931076CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/289310762025-06-30T13:40:07Z
spellingShingle Cellwise Outlier Detection in Heterogeneous Populations
Giorgia Zaccaria (21245577)
Biotechnology
Ecology
Cancer
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
Cellwise contamination
EM algorithm
Imputation
Missing data
Model-based clustering
Robustness
status_str publishedVersion
title Cellwise Outlier Detection in Heterogeneous Populations
title_full Cellwise Outlier Detection in Heterogeneous Populations
title_fullStr Cellwise Outlier Detection in Heterogeneous Populations
title_full_unstemmed Cellwise Outlier Detection in Heterogeneous Populations
title_short Cellwise Outlier Detection in Heterogeneous Populations
title_sort Cellwise Outlier Detection in Heterogeneous Populations
topic Biotechnology
Ecology
Cancer
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
Cellwise contamination
EM algorithm
Imputation
Missing data
Model-based clustering
Robustness