1. Identity statement | |
Reference Type | Conference Paper (Conference Proceedings) |
Site | mtc-m21d.sid.inpe.br |
Holder Code | isadg {BR SPINPE} ibi 8JMKD3MGPCW/3DT298S |
Identifier | 8JMKD3MGP3W34T/45PLTT8 |
Repository | sid.inpe.br/mtc-m21d/2021/11.11.18.46 |
Last Update | 2023:12.28.17.43.49 (UTC) simone |
Metadata Repository | sid.inpe.br/mtc-m21d/2021/11.11.18.46.52 |
Metadata Last Update | 2023:12.28.17.43.51 (UTC) simone |
Secondary Key | INPE--PRE/ |
Citation Key | Garcia:2021:ExTéAp |
Title | Explorando técnicas de aprendizado de máquina para aprimoramento da previsão de geadas no Sul e Sudeste do Brasil |
Year | 2021 |
Access Date | 2024, May 19 |
Secondary Type | PRE CN |
Number of Files | 1 |
Size | 5925 KiB |
|
2. Context | |
Author | Garcia, José Roberto Motta |
Resume Identifier | 8JMKD3MGP5W/3C9JHJD |
Group | DIMNT-CGCT-INPE-MCTI-GOV-BR |
Affiliation | Instituto Nacional de Pesquisas Espaciais (INPE) |
Author e-Mail Address | garcia.cptec@gmail.com |
Conference Name | Congresso Brasileiro de Agroinformática (SIBIAGRO), 13 |
Date | 10-12 nov. |
History (UTC) | 2021-11-11 18:47:02 :: simone -> administrator :: 2021 2022-04-03 22:29:02 :: administrator -> simone :: 2021 |
|
3. Content and structure | |
Is the master or a copy? | is the master |
Content Stage | completed |
Transferable | 1 |
Content Type | External Contribution |
Version Type | publisher |
Abstract | The CPTEC/INPE operational frost forecast is based on an index established from the weighted variation of meteorological variables from a weather forecasting model, which is compared to the operationally predicted values. Constant improvement of these forecasts is necessary to mitigate the economic and social damage caused by frosts. The objective of this work is to identify a methodology based on machine learning that surpasses the current prediction system. Therefore, this exploratory research applied different combinations between three algorithms, varying their parameterizations and the set of predictor variables from meteorological station data in southern and southeastern Brazil. Preliminary tests on 24-hour forecasts show that performance surpasses the current method, signaling that the search for the most suitable methodology should be continued. Resumo. A previsão operacional de geadas do CPTEC/INPE é baseada em um índice estabelecido a partir da variação ponderada das variáveis meteorológicas de um modelo de previsão de tempo, que é comparado aos valores previstos operacionalmente. Um constante aprimoramento dessas previsões visando mitigar danos econômicos e sociais decorrentes de geadas se faz necessária. O propósito deste trabalho é identificar uma metodologia baseada em aprendizado de máquina que supere o desempenho do sistema de previsão atual. Para tanto, esta pesquisa exploratória aplicou diferentes combinações entre três algoritmos, variando suas parametrizações e o conjunto de variáveis preditoras de dados de estações meteorológicas do sul e sudeste do Brasil. Testes preliminares em previsões de 24h mostram que o desempenho supera o método atual, sinalizando que o processo de busca da metodologia mais adequada deve ser continuada. |
Area | MET |
Arrangement | urlib.net > BDMCI > Fonds > Produção a partir de 2021 > CGCT > Explorando técnicas de... |
doc Directory Content | access |
source Directory Content | there are no files |
agreement Directory Content | |
|
4. Conditions of access and use | |
data URL | http://urlib.net/ibi/8JMKD3MGP3W34T/45PLTT8 |
zipped data URL | http://urlib.net/zip/8JMKD3MGP3W34T/45PLTT8 |
Language | pt |
Target File | garcia_explorando.pdf |
User Group | simone |
Reader Group | administrator simone |
Visibility | shown |
Update Permission | not transferred |
|
5. Allied materials | |
Next Higher Units | 8JMKD3MGPCW/46KUATE |
Citing Item List | sid.inpe.br/mtc-m21/2012/07.13.14.52.32 1 |
Host Collection | urlib.net/www/2021/06.04.03.40 |
|
6. Notes | |
Empty Fields | archivingpolicy archivist booktitle callnumber conferencelocation copyholder copyright creatorhistory descriptionlevel dissemination doi e-mailaddress edition editor format isbn issn keywords label lineage mark mirrorrepository nextedition notes numberofvolumes orcid organization pages parameterlist parentrepositories previousedition previouslowerunit progress project publisher publisheraddress readpermission rightsholder schedulinginformation secondarydate secondarymark serieseditor session shorttitle sponsor subject tertiarymark tertiarytype type url volume |
|
7. Description control | |
e-Mail (login) | simone |
update | |
|