Application of Entropy Ensemble Filter in Neural Network Forecasts of Tropical Pacific Sea Surface Temperatures
Recently, the Entropy Ensemble Filter (EEF) method was proposed to mitigate the computational cost of the Bootstrap AGGregatING (bagging) method. This method uses the most informative training data sets in the model ensemble rather than all ensemble members created by the conventional bagging. In th...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2018-03-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | http://www.mdpi.com/1099-4300/20/3/207 |
| _version_ | 1811279285818753024 |
|---|---|
| author | Hossein Foroozand Valentina Radić Steven V. Weijs |
| author_facet | Hossein Foroozand Valentina Radić Steven V. Weijs |
| author_sort | Hossein Foroozand |
| collection | DOAJ |
| description | Recently, the Entropy Ensemble Filter (EEF) method was proposed to mitigate the computational cost of the Bootstrap AGGregatING (bagging) method. This method uses the most informative training data sets in the model ensemble rather than all ensemble members created by the conventional bagging. In this study, we evaluate, for the first time, the application of the EEF method in Neural Network (NN) modeling of El Nino-southern oscillation. Specifically, we forecast the first five principal components (PCs) of sea surface temperature monthly anomaly fields over tropical Pacific, at different lead times (from 3 to 15 months, with a three-month increment) for the period 1979–2017. We apply the EEF method in a multiple-linear regression (MLR) model and two NN models, one using Bayesian regularization and one Levenberg-Marquardt algorithm for training, and evaluate their performance and computational efficiency relative to the same models with conventional bagging. All models perform equally well at the lead time of 3 and 6 months, while at higher lead times, the MLR model’s skill deteriorates faster than the nonlinear models. The neural network models with both bagging methods produce equally successful forecasts with the same computational efficiency. It remains to be shown whether this finding is sensitive to the dataset size. |
| first_indexed | 2024-04-13T00:51:55Z |
| format | Article |
| id | doaj.art-3e3cc668785d463ba19d62151f246351 |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-04-13T00:51:55Z |
| publishDate | 2018-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-3e3cc668785d463ba19d62151f2463512022-12-22T03:09:51ZengMDPI AGEntropy1099-43002018-03-0120320710.3390/e20030207e20030207Application of Entropy Ensemble Filter in Neural Network Forecasts of Tropical Pacific Sea Surface TemperaturesHossein Foroozand0Valentina Radić1Steven V. Weijs2Department of Civil Engineering, University of British Columbia, Vancouver, BC V6T 1Z4, CanadaDepartment of Earth, Ocean and Atmospheric Sciences, University of British Columbia, Vancouver, BC V6T 1Z4, CanadaDepartment of Civil Engineering, University of British Columbia, Vancouver, BC V6T 1Z4, CanadaRecently, the Entropy Ensemble Filter (EEF) method was proposed to mitigate the computational cost of the Bootstrap AGGregatING (bagging) method. This method uses the most informative training data sets in the model ensemble rather than all ensemble members created by the conventional bagging. In this study, we evaluate, for the first time, the application of the EEF method in Neural Network (NN) modeling of El Nino-southern oscillation. Specifically, we forecast the first five principal components (PCs) of sea surface temperature monthly anomaly fields over tropical Pacific, at different lead times (from 3 to 15 months, with a three-month increment) for the period 1979–2017. We apply the EEF method in a multiple-linear regression (MLR) model and two NN models, one using Bayesian regularization and one Levenberg-Marquardt algorithm for training, and evaluate their performance and computational efficiency relative to the same models with conventional bagging. All models perform equally well at the lead time of 3 and 6 months, while at higher lead times, the MLR model’s skill deteriorates faster than the nonlinear models. The neural network models with both bagging methods produce equally successful forecasts with the same computational efficiency. It remains to be shown whether this finding is sensitive to the dataset size.http://www.mdpi.com/1099-4300/20/3/207entropy ensemble filterensemble model simulation criterionEEF methodbootstrap aggregatingbaggingbootstrap neural networksEl NiñoENSOneural network forecastsea surface temperaturetropical Pacific |
| spellingShingle | Hossein Foroozand Valentina Radić Steven V. Weijs Application of Entropy Ensemble Filter in Neural Network Forecasts of Tropical Pacific Sea Surface Temperatures Entropy entropy ensemble filter ensemble model simulation criterion EEF method bootstrap aggregating bagging bootstrap neural networks El Niño ENSO neural network forecast sea surface temperature tropical Pacific |
| title | Application of Entropy Ensemble Filter in Neural Network Forecasts of Tropical Pacific Sea Surface Temperatures |
| title_full | Application of Entropy Ensemble Filter in Neural Network Forecasts of Tropical Pacific Sea Surface Temperatures |
| title_fullStr | Application of Entropy Ensemble Filter in Neural Network Forecasts of Tropical Pacific Sea Surface Temperatures |
| title_full_unstemmed | Application of Entropy Ensemble Filter in Neural Network Forecasts of Tropical Pacific Sea Surface Temperatures |
| title_short | Application of Entropy Ensemble Filter in Neural Network Forecasts of Tropical Pacific Sea Surface Temperatures |
| title_sort | application of entropy ensemble filter in neural network forecasts of tropical pacific sea surface temperatures |
| topic | entropy ensemble filter ensemble model simulation criterion EEF method bootstrap aggregating bagging bootstrap neural networks El Niño ENSO neural network forecast sea surface temperature tropical Pacific |
| url | http://www.mdpi.com/1099-4300/20/3/207 |
| work_keys_str_mv | AT hosseinforoozand applicationofentropyensemblefilterinneuralnetworkforecastsoftropicalpacificseasurfacetemperatures AT valentinaradic applicationofentropyensemblefilterinneuralnetworkforecastsoftropicalpacificseasurfacetemperatures AT stevenvweijs applicationofentropyensemblefilterinneuralnetworkforecastsoftropicalpacificseasurfacetemperatures |