Logo do repositório
 
Publicação

An Ensemble Learning Strategy for Panel Time Series Forecasting of Excess Mortality During the COVID-19 Pandemic

2022-10-01Artigo científico Acesso aberto
Ver registo simples
dc.contributor.authorAshofteh, Afshin
dc.contributor.authorBravo, Jorge Miguel
dc.contributor.authorAyuso, Mercedes
dc.contributor.institutionNOVA Information Management School (NOVA IMS)
dc.contributor.institutionInformation Management Research Center (MagIC) - NOVA Information Management School
dc.contributor.pblElsevier Science B.V., Amsterdam.
dc.date.accessioned2022-05-25T22:18:17Z
dc.date.available2022-05-25T22:18:17Z
dc.date.issued2022-10-01
dc.descriptionAshofteh, A., Bravo, J. M., & Ayuso, M. (2022). An Ensemble Learning Strategy for Panel Time Series Forecasting of Excess Mortality During the COVID-19 Pandemic. Applied Soft Computing, 128(October), 1-17. [109422]. https://doi.org/10.2139/ssrn.4057314, https://doi.org/10.1016/j.asoc.2022.109422 ----- Fundinhg: The authors are grateful to the anonymous reviewers for their constructive comments. Afshin Ashofteh and Jorge M. Bravo were supported by Portuguese national science funds made available through the FCT under project UIDB/04152/2020-Centro de Investigação em Gestão de Informação (MagIC). Additionally, Mercedes Ayuso is grateful to the Spanish Ministry of Science and Innovation for funding received under grant PID2019-105986GB-C21 and to the Secretaria d’Universitats i Recerca del departament d’Empresa i Coneixement de la Generalitat de Catalunya for funding received under grant 2020-PANDE-00074 (research project directly related to COVID-19 and economy).
dc.description.abstractQuantifying and analyzing excess mortality in crises such as the ongoing COVID-19 pandemic is crucial for policymakers. The traditional way it is measured does not account for differences in the level, long-term secular trends, and seasonal patterns in all-cause mortality across countries and regions. This paper develops and empirically investigates the forecasting performance of a novel flexible and dynamic ensemble learning strategy for seasonal time series forecasting of monthly respiratory diseases deaths data across a pool of 61 heterogeneous countries. The strategy is based on a Bayesian Model Ensemble (BME) of heterogeneous time series methods involving both the selection of the subset of best forecasters (model confidence set), the identification of the best holdout period for each contributed model, and the determination of optimal weights using the out-of-sample predictive accuracy. A model selection strategy is also developed to remove the outlier models and to combine the models with reasonable accuracy in the ensemble. The empirical results of this large set of experiments show that the accuracy of the BME approach improves noticeably by using a flexible and dynamic holdout period selection. Additionally, that the BME forecasts of respiratory disease deaths for each country are highly accurate and exhibit a high correlation (94%) with COVID-19 deaths in 2020.en
dc.description.versionpublishersversion
dc.description.versionpublished
dc.format.extent17
dc.format.extent3034856
dc.identifier.doi10.1016/j.asoc.2022.109422
dc.identifier.issn1568-4946
dc.identifier.otherPURE: 44265559
dc.identifier.otherPURE UUID: 2f2c39c2-cd33-4b93-b151-a15dd23a46c0
dc.identifier.othercrossref: 10.2139/ssrn.4057314
dc.identifier.otherScopus: 85136615085
dc.identifier.otherWOS: 000865430700005
dc.identifier.otherORCID: /0000-0002-7389-5103/work/116782060
dc.identifier.urihttp://hdl.handle.net/10362/138671
dc.identifier.urlhttps://www.scopus.com/pages/publications/85136615085
dc.identifier.urlhttps://www.webofscience.com/wos/woscc/full-record/WOS:000865430700005
dc.identifier.urlhttps://data.mendeley.com/datasets/gjj68bmv8d/2
dc.identifier.urlhttps://www.ssrn.com/abstract=4057314
dc.language.isoeng
dc.peerreviewedyes
dc.relationinfo:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDB%2F04152%2F2020/PT
dc.relationInformation Management Research Center
dc.subjectLayered learning
dc.subjectMultiple learning processes
dc.subjectTime Series
dc.subjectEnsemble Bayesian Model Averaging (EBMA)
dc.subjectSARS-CoV-2
dc.subjectCOVID-19
dc.subjectBayesian model averaging (BMA)
dc.subjectEnsemble learning
dc.subjectForecasting
dc.subjectPanel data
dc.subjectMachine learning
dc.subjectSoftware
dc.subjectSDG 3 - Good Health and Well-being
dc.titleAn Ensemble Learning Strategy for Panel Time Series Forecasting of Excess Mortality During the COVID-19 Pandemicen
dc.typejournal article
degois.publication.firstPage1
degois.publication.issueOctober
degois.publication.lastPage17
degois.publication.titleApplied Soft Computing
degois.publication.volume128
dspace.entity.typePublication
oaire.awardNumberUIDB/04152/2020
oaire.awardTitleInformation Management Research Center
oaire.awardURIinfo:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDB%2F04152%2F2020/PT
oaire.fundingStream6817 - DCRRNI ID
project.funder.identifierhttp://doi.org/10.13039/501100001871
project.funder.nameFundação para a Ciência e a Tecnologia
rcaap.rightsopenAccess
relation.isProjectOfPublication3274bdb3-4dd3-4bbe-8f74-d34190081f87
relation.isProjectOfPublication.latestForDiscovery3274bdb3-4dd3-4bbe-8f74-d34190081f87

Ficheiros

Principais
A mostrar 1 - 1 de 1
Miniatura
Nome:
1_s2.0_S156849462200552X_main.pdf
Tamanho:
2.89 MB
Formato:
Adobe Portable Document Format
Ver/Abrir

Coleções

NIMS: MagIC - Artigos em revista internacional com arbitragem científica (Peer-Review articles in international journals)