Nowcasting brazilian inflation with machine learning

Detalhes bibliográficos
Autor(a) principal: Garnitskiy, Leonid
Data de Publicação: 2020
Tipo de documento: Dissertação
Idioma: eng
Título da fonte: Repositório Institucional do FGV (FGV Repositório Digital)
Texto Completo: https://hdl.handle.net/10438/31426
Resumo: A importância de monitoramento das principais variáveis macroeconômicas é evidenciada pelo grande esforço que os agentes devotam a esta tarefa. O presente trabalho propõe-se a contribuir para a literatura de previsão econômica, aplicando os modelos de aprendizado de máquina para monitorar diariamente a inflação brasileira medida pelo IPCA. Os resultados obtidos são promissores. O benefício de fazer monitoramento diário da inflação em vez da previsão uma vez por mês é na ordem de 50%-60% em média para quase todos os modelos de aprendizado de máquina considerados. Os modelos que apresentam o melhor desempenho são Regressão de Subconjunto Completo e Floresta Aleatória. Os resultados também mostram que usar técnicas multivariadas de aprendizado de máquina em vez de simples modelos univariados reduz o erro da previsão em até 20%.
id FGV_011b4c3b13f3356e810b09c3dd993310
oai_identifier_str oai:repositorio.fgv.br:10438/31426
network_acronym_str FGV
network_name_str Repositório Institucional do FGV (FGV Repositório Digital)
repository_id_str 3974
spelling Garnitskiy, LeonidEscolas::EPGEIachan, Felipe SaraivaGaglianone, Wagner PiazzaIssler, João Victor2021-12-17T17:37:30Z2021-12-17T17:37:30Z2020-04-24https://hdl.handle.net/10438/31426A importância de monitoramento das principais variáveis macroeconômicas é evidenciada pelo grande esforço que os agentes devotam a esta tarefa. O presente trabalho propõe-se a contribuir para a literatura de previsão econômica, aplicando os modelos de aprendizado de máquina para monitorar diariamente a inflação brasileira medida pelo IPCA. Os resultados obtidos são promissores. O benefício de fazer monitoramento diário da inflação em vez da previsão uma vez por mês é na ordem de 50%-60% em média para quase todos os modelos de aprendizado de máquina considerados. Os modelos que apresentam o melhor desempenho são Regressão de Subconjunto Completo e Floresta Aleatória. Os resultados também mostram que usar técnicas multivariadas de aprendizado de máquina em vez de simples modelos univariados reduz o erro da previsão em até 20%.Many efforts are made by economists to track key macroeconomic variables in real time. This paper aims to make a contribution to economic forecasting research by employing Machine Learning techniques to perform a daily nowcasting of Brazilian inflation. The original results obtained are encouraging. The benefit of making a daily nowcast of inflation instead of one-month-ahead forecast is found to be of 50%-60% on average for almost all ML models considered. The best-performing ML techniques are Complete Subset Regression and Random Forest. The results also show that using ML methods instead of univariate benchmarks reduces the nowcasting error in at most 20%.engMonitoramento da inflaçãoAprendizado de máquinaRegressão de subconjunto completoInflation nowcastingMachine learningComplete subset regressionsEconomiaInflação - Brasil - Modelos macroeconômicosAprendizado do computadorModelagem de dadosNowcasting brazilian inflation with machine learninginfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesis2020-04-24info:eu-repo/semantics/openAccessreponame:Repositório Institucional do FGV (FGV Repositório Digital)instname:Fundação Getulio Vargas (FGV)instacron:FGVORIGINALPDFPDFapplication/pdf4660319https://repositorio.fgv.br/bitstreams/39c7346b-46ab-4c2c-8dd6-4cad17b1a1ee/download75851c6c9afadf77ec6799a4401a0488MD51LICENSElicense.txtlicense.txttext/plain; charset=utf-84707https://repositorio.fgv.br/bitstreams/3fee828a-26bd-412f-ad0a-24941e7c928c/downloaddfb340242cced38a6cca06c627998fa1MD52TEXTDISSERTACAO.pdf.txtDISSERTACAO.pdf.txtExtracted texttext/plain102318https://repositorio.fgv.br/bitstreams/e54d03d0-bd11-4856-9e63-d879c673f38a/download1c55f8d3d0d08935aa1a8fba310dd626MD55PDF.txtPDF.txtExtracted texttext/plain102318https://repositorio.fgv.br/bitstreams/3f702535-00c5-4e8c-a1ff-5820ae689efd/download1c55f8d3d0d08935aa1a8fba310dd626MD57THUMBNAILDISSERTACAO.pdf.jpgDISSERTACAO.pdf.jpgGenerated Thumbnailimage/jpeg2735https://repositorio.fgv.br/bitstreams/2fd6f10d-0919-43e4-acb6-bdf485fe9827/downloadddb35f3d9d07a4886a5768749bff2f02MD56PDF.jpgPDF.jpgGenerated Thumbnailimage/jpeg2735https://repositorio.fgv.br/bitstreams/1cdd9e48-ac65-4a98-abac-280212bd1af4/downloadddb35f3d9d07a4886a5768749bff2f02MD5810438/314262024-07-08 19:17:38.94open.accessoai:repositorio.fgv.br:10438/31426https://repositorio.fgv.brRepositório InstitucionalPRIhttp://bibliotecadigital.fgv.br/dspace-oai/requestopendoar:39742024-07-08T19:17:38Repositório Institucional do FGV (FGV Repositório Digital) - Fundação Getulio Vargas (FGV)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
dc.title.eng.fl_str_mv Nowcasting brazilian inflation with machine learning
title Nowcasting brazilian inflation with machine learning
spellingShingle Nowcasting brazilian inflation with machine learning
Garnitskiy, Leonid
Monitoramento da inflação
Aprendizado de máquina
Regressão de subconjunto completo
Inflation nowcasting
Machine learning
Complete subset regressions
Economia
Inflação - Brasil - Modelos macroeconômicos
Aprendizado do computador
Modelagem de dados
title_short Nowcasting brazilian inflation with machine learning
title_full Nowcasting brazilian inflation with machine learning
title_fullStr Nowcasting brazilian inflation with machine learning
title_full_unstemmed Nowcasting brazilian inflation with machine learning
title_sort Nowcasting brazilian inflation with machine learning
author Garnitskiy, Leonid
author_facet Garnitskiy, Leonid
author_role author
dc.contributor.unidadefgv.por.fl_str_mv Escolas::EPGE
dc.contributor.member.none.fl_str_mv Iachan, Felipe Saraiva
Gaglianone, Wagner Piazza
dc.contributor.author.fl_str_mv Garnitskiy, Leonid
dc.contributor.advisor1.fl_str_mv Issler, João Victor
contributor_str_mv Issler, João Victor
dc.subject.por.fl_str_mv Monitoramento da inflação
Aprendizado de máquina
Regressão de subconjunto completo
topic Monitoramento da inflação
Aprendizado de máquina
Regressão de subconjunto completo
Inflation nowcasting
Machine learning
Complete subset regressions
Economia
Inflação - Brasil - Modelos macroeconômicos
Aprendizado do computador
Modelagem de dados
dc.subject.eng.fl_str_mv Inflation nowcasting
Machine learning
Complete subset regressions
dc.subject.area.por.fl_str_mv Economia
dc.subject.bibliodata.por.fl_str_mv Inflação - Brasil - Modelos macroeconômicos
Aprendizado do computador
Modelagem de dados
description A importância de monitoramento das principais variáveis macroeconômicas é evidenciada pelo grande esforço que os agentes devotam a esta tarefa. O presente trabalho propõe-se a contribuir para a literatura de previsão econômica, aplicando os modelos de aprendizado de máquina para monitorar diariamente a inflação brasileira medida pelo IPCA. Os resultados obtidos são promissores. O benefício de fazer monitoramento diário da inflação em vez da previsão uma vez por mês é na ordem de 50%-60% em média para quase todos os modelos de aprendizado de máquina considerados. Os modelos que apresentam o melhor desempenho são Regressão de Subconjunto Completo e Floresta Aleatória. Os resultados também mostram que usar técnicas multivariadas de aprendizado de máquina em vez de simples modelos univariados reduz o erro da previsão em até 20%.
publishDate 2020
dc.date.issued.fl_str_mv 2020-04-24
dc.date.accessioned.fl_str_mv 2021-12-17T17:37:30Z
dc.date.available.fl_str_mv 2021-12-17T17:37:30Z
dc.type.status.fl_str_mv info:eu-repo/semantics/publishedVersion
dc.type.driver.fl_str_mv info:eu-repo/semantics/masterThesis
format masterThesis
status_str publishedVersion
dc.identifier.uri.fl_str_mv https://hdl.handle.net/10438/31426
url https://hdl.handle.net/10438/31426
dc.language.iso.fl_str_mv eng
language eng
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.source.none.fl_str_mv reponame:Repositório Institucional do FGV (FGV Repositório Digital)
instname:Fundação Getulio Vargas (FGV)
instacron:FGV
instname_str Fundação Getulio Vargas (FGV)
instacron_str FGV
institution FGV
reponame_str Repositório Institucional do FGV (FGV Repositório Digital)
collection Repositório Institucional do FGV (FGV Repositório Digital)
bitstream.url.fl_str_mv https://repositorio.fgv.br/bitstreams/39c7346b-46ab-4c2c-8dd6-4cad17b1a1ee/download
https://repositorio.fgv.br/bitstreams/3fee828a-26bd-412f-ad0a-24941e7c928c/download
https://repositorio.fgv.br/bitstreams/e54d03d0-bd11-4856-9e63-d879c673f38a/download
https://repositorio.fgv.br/bitstreams/3f702535-00c5-4e8c-a1ff-5820ae689efd/download
https://repositorio.fgv.br/bitstreams/2fd6f10d-0919-43e4-acb6-bdf485fe9827/download
https://repositorio.fgv.br/bitstreams/1cdd9e48-ac65-4a98-abac-280212bd1af4/download
bitstream.checksum.fl_str_mv 75851c6c9afadf77ec6799a4401a0488
dfb340242cced38a6cca06c627998fa1
1c55f8d3d0d08935aa1a8fba310dd626
1c55f8d3d0d08935aa1a8fba310dd626
ddb35f3d9d07a4886a5768749bff2f02
ddb35f3d9d07a4886a5768749bff2f02
bitstream.checksumAlgorithm.fl_str_mv MD5
MD5
MD5
MD5
MD5
MD5
repository.name.fl_str_mv Repositório Institucional do FGV (FGV Repositório Digital) - Fundação Getulio Vargas (FGV)
repository.mail.fl_str_mv
_version_ 1813797854051827712