| |
|
|
| Registros recuperados : 926 | |
| Registros recuperados : 926 | |
|
|
|
 | Acesso ao texto completo restrito à biblioteca da Embrapa Cerrados. Para informações adicionais entre em contato com cpac.biblioteca@embrapa.br. |
Registro Completo
|
Biblioteca(s): |
Embrapa Cerrados. |
|
Data corrente: |
19/08/2021 |
|
Data da última atualização: |
19/08/2021 |
|
Tipo da produção científica: |
Artigo em Periódico Indexado |
|
Circulação/Nível: |
A - 1 |
|
Autoria: |
ALTHOFF, D.; RODRIGUES, L. N.; BAZAME, H. C. |
|
Afiliação: |
DANIEL ALTHOFF; LINEU NEIVA RODRIGUES, CPAC; HELIZANI COUTO BAZAME. |
|
Título: |
Uncertainty quantification for hydrological models based on neural networks: the dropout ensemble. |
|
Ano de publicação: |
2021 |
|
Fonte/Imprenta: |
Stochastic Environmental Research and Risk Assessment, v. 35, p. 1051?1067, 2021. |
|
Páginas: |
p. 1051-1067 |
|
DOI: |
https://doi.org/10.1007/s00477-021-01980-8 |
|
Idioma: |
Inglês |
|
Conteúdo: |
Abstract The use of neural networks in hydrology has been frequently undermined by limitations regarding the quantification of uncertainty in predictions. Many authors have proposed different methodologies to overcome these limitations, such as running Monte Carlo simulations, Bayesian approximations, and bootstrapping training samples, which come with computational limitations of their own, and two-step approaches, among others. One less-frequently explored alternative is to repurpose the dropout scheme during inference. Dropout is commonly used during training to avoid overfitting. However, it may also be activated during the testing period to effortlessly provide an ensemble of multiple ??sister?? predictions. This study explores the predictive uncertainty in hydrological models based on neural networks by comparing a multiparameter ensemble to a dropout ensemble. The dropout ensemble shows more reliable coverage of prediction intervals, while the multiparameter ensemble results in sharper prediction intervals. Moreover, for neural network structures with optimal lookback series, both ensemble strategies result in similar average interval scores. The dropout ensemble, however, benefits from requiring only a single calibration run, i.e., a single set of parameters. In addition, it delivers important insight for engineering design and decision-making with no increase in computational cost. Therefore, the dropout ensemble can be easily included in uncertainty analysis routines and even be combined with multiparameter or multimodel alternatives. MenosAbstract The use of neural networks in hydrology has been frequently undermined by limitations regarding the quantification of uncertainty in predictions. Many authors have proposed different methodologies to overcome these limitations, such as running Monte Carlo simulations, Bayesian approximations, and bootstrapping training samples, which come with computational limitations of their own, and two-step approaches, among others. One less-frequently explored alternative is to repurpose the dropout scheme during inference. Dropout is commonly used during training to avoid overfitting. However, it may also be activated during the testing period to effortlessly provide an ensemble of multiple ??sister?? predictions. This study explores the predictive uncertainty in hydrological models based on neural networks by comparing a multiparameter ensemble to a dropout ensemble. The dropout ensemble shows more reliable coverage of prediction intervals, while the multiparameter ensemble results in sharper prediction intervals. Moreover, for neural network structures with optimal lookback series, both ensemble strategies result in similar average interval scores. The dropout ensemble, however, benefits from requiring only a single calibration run, i.e., a single set of parameters. In addition, it delivers important insight for engineering design and decision-making with no increase in computational cost. Therefore, the dropout ensemble can be easily included in uncertainty analysis routin... Mostrar Tudo |
|
Palavras-Chave: |
Modelo hidrológico; Rede neural. |
|
Thesagro: |
Hidrologia. |
|
Categoria do assunto: |
-- |
|
Marc: |
LEADER 02241naa a2200205 a 4500
001 2133739
005 2021-08-19
008 2021 bl uuuu u00u1 u #d
024 7 $ahttps://doi.org/10.1007/s00477-021-01980-8$2DOI
100 1 $aALTHOFF, D.
245 $aUncertainty quantification for hydrological models based on neural networks$bthe dropout ensemble.$h[electronic resource]
260 $c2021
300 $ap. 1051-1067
520 $aAbstract The use of neural networks in hydrology has been frequently undermined by limitations regarding the quantification of uncertainty in predictions. Many authors have proposed different methodologies to overcome these limitations, such as running Monte Carlo simulations, Bayesian approximations, and bootstrapping training samples, which come with computational limitations of their own, and two-step approaches, among others. One less-frequently explored alternative is to repurpose the dropout scheme during inference. Dropout is commonly used during training to avoid overfitting. However, it may also be activated during the testing period to effortlessly provide an ensemble of multiple ??sister?? predictions. This study explores the predictive uncertainty in hydrological models based on neural networks by comparing a multiparameter ensemble to a dropout ensemble. The dropout ensemble shows more reliable coverage of prediction intervals, while the multiparameter ensemble results in sharper prediction intervals. Moreover, for neural network structures with optimal lookback series, both ensemble strategies result in similar average interval scores. The dropout ensemble, however, benefits from requiring only a single calibration run, i.e., a single set of parameters. In addition, it delivers important insight for engineering design and decision-making with no increase in computational cost. Therefore, the dropout ensemble can be easily included in uncertainty analysis routines and even be combined with multiparameter or multimodel alternatives.
650 $aHidrologia
653 $aModelo hidrológico
653 $aRede neural
700 1 $aRODRIGUES, L. N.
700 1 $aBAZAME, H. C.
773 $tStochastic Environmental Research and Risk Assessment$gv. 35, p. 1051?1067, 2021.
Download
Esconder MarcMostrar Marc Completo |
|
Registro original: |
Embrapa Cerrados (CPAC) |
|
|
Biblioteca |
ID |
Origem |
Tipo/Formato |
Classificação |
Cutter |
Registro |
Volume |
Status |
Fechar
|
| Nenhum registro encontrado para a expressão de busca informada. |
|
|
