Exposome data drift: implications for machine learning based diabetes prediction

Treballs finals del Màster de Fonaments de Ciència de Dades, Facultat de matemàtiques, Universitat de Barcelona. Curs: 2022-2023. Tutor: Karim Lekadir i Marina Camacho

Detalles Bibliográficos
Autor: Brosten, Peter Hannagan
Tipo de recurso: tesis de maestría
Fecha de publicación:2023
País:España
Institución:Universidad de Barcelona
Repositorio:Dipòsit Digital de la UB
OAI Identifier:oai:diposit.ub.edu:2445/212913
Acceso en línea:https://hdl.handle.net/2445/212913
Access Level:acceso abierto
Palabra clave:Aprenentatge automàtic
Diabetis
Dades massives
Treballs de fi de màster
Machine learning
Diabetes
Big data
Master's thesis
id ES_144d4e8a9221e99f6a554ebe0b821118
oai_identifier_str oai:diposit.ub.edu:2445/212913
network_acronym_str ES
network_name_str España
repository_id_str
spelling Exposome data drift: implications for machine learning based diabetes predictionBrosten, Peter HannaganAprenentatge automàticDiabetisDades massivesTreballs de fi de màsterMachine learningDiabetesBig dataMaster's thesisTreballs finals del Màster de Fonaments de Ciència de Dades, Facultat de matemàtiques, Universitat de Barcelona. Curs: 2022-2023. Tutor: Karim Lekadir i Marina Camacho[en] Data drift is a problem in machine learning (ML) where characteristics of the input predictors changes over time, leading to model degradation. However, the effects of data drift on ML models built from human exposome data have not been well described yet. This study aimed to investigate data drifts for exposome data in ML models of diabetes risk. 7,521 participants with a diagnosis of diabetes from the UK Biobank, along with a proportional control group from 2006 to 2010 were used to train several baseline ML models for diabetes prediction. A second cohort of 4,007 participants attending the follow-up assessment period from 2012 to 2013 was used to assess potential data drifts over time. When evaluated on the second cohort, significant performance degradation was found in all baseline models (i.e.average precision dropped by 15%, f1-score by 12%, recall by 15%, and precision by 10%). A suite of drift detection tests were run on the best performing baseline models to identify possible signatures of three distinct kinds of data drift: covariate drift, label drift, and concept drift. Utilizing both multivariate and univariate data distribution based detection methods, covariate drift was identified in features such as Birth Year, BMI, Frequency of Tiredness, and Lack of Education. A comparison of prevalence rates for time-ordered batches of the population found no severe label drift. Nonetheless, gradual label drift could not be ruled out. A model-aware concept drift detection method was employed, monitoring temporal changes in normalized Shapley contributions for the model’s input features. This test found drift in abnormal changes in feature contribution when predicting on the second cohort for the Birth Year feature and near alerts in multiple others. This study shows the potential for data drift acting as a driver of model degradation in exposome-based ML models and highlights the need for further research into the traceability of clinical AI/ML solutions.Lekadir, Karim, 1977-Camacho, Marina2023info:eu-repo/semantics/masterThesisapplication/pdfhttps://hdl.handle.net/2445/212913Màster Oficial - Fonaments de la Ciència de Dadesreponame:Dipòsit Digital de la UBinstname:Universidad de BarcelonaIngléscc-by-nc-nd (c) Peter Hannagan Brosten, 2023http://creativecommons.org/licenses/by-nc-nd/3.0/es/info:eu-repo/semantics/openAccessoai:diposit.ub.edu:2445/2129132026-05-27T06:46:51Z
dc.title.none.fl_str_mv Exposome data drift: implications for machine learning based diabetes prediction
title Exposome data drift: implications for machine learning based diabetes prediction
spellingShingle Exposome data drift: implications for machine learning based diabetes prediction
Brosten, Peter Hannagan
Aprenentatge automàtic
Diabetis
Dades massives
Treballs de fi de màster
Machine learning
Diabetes
Big data
Master's thesis
title_short Exposome data drift: implications for machine learning based diabetes prediction
title_full Exposome data drift: implications for machine learning based diabetes prediction
title_fullStr Exposome data drift: implications for machine learning based diabetes prediction
title_full_unstemmed Exposome data drift: implications for machine learning based diabetes prediction
title_sort Exposome data drift: implications for machine learning based diabetes prediction
dc.creator.none.fl_str_mv Brosten, Peter Hannagan
author Brosten, Peter Hannagan
author_facet Brosten, Peter Hannagan
author_role author
dc.contributor.none.fl_str_mv Lekadir, Karim, 1977-
Camacho, Marina
dc.subject.none.fl_str_mv Aprenentatge automàtic
Diabetis
Dades massives
Treballs de fi de màster
Machine learning
Diabetes
Big data
Master's thesis
topic Aprenentatge automàtic
Diabetis
Dades massives
Treballs de fi de màster
Machine learning
Diabetes
Big data
Master's thesis
description Treballs finals del Màster de Fonaments de Ciència de Dades, Facultat de matemàtiques, Universitat de Barcelona. Curs: 2022-2023. Tutor: Karim Lekadir i Marina Camacho
publishDate 2023
dc.date.none.fl_str_mv 2023
dc.type.none.fl_str_mv info:eu-repo/semantics/masterThesis
format masterThesis
dc.identifier.none.fl_str_mv https://hdl.handle.net/2445/212913
url https://hdl.handle.net/2445/212913
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.rights.none.fl_str_mv cc-by-nc-nd (c) Peter Hannagan Brosten, 2023
http://creativecommons.org/licenses/by-nc-nd/3.0/es/
info:eu-repo/semantics/openAccess
rights_invalid_str_mv cc-by-nc-nd (c) Peter Hannagan Brosten, 2023
http://creativecommons.org/licenses/by-nc-nd/3.0/es/
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.source.none.fl_str_mv Màster Oficial - Fonaments de la Ciència de Dades
reponame:Dipòsit Digital de la UB
instname:Universidad de Barcelona
instname_str Universidad de Barcelona
reponame_str Dipòsit Digital de la UB
collection Dipòsit Digital de la UB
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869403728950329344
score 15,811543

Similares en LA Referencia