A hybrid optimized error correction system for time series forecasting

Oliveira, João Fausto Lorenzato de; Pacífico, Luciano D. S.; Neto, Paulo S. G. de Mattos; Barreiros, Emanoel; Rodrigues, Cleyton Mário de Oliveira; Filho, Adauto Trigueiro de Almeida

doi:10.1016/j.asoc.2019.105970

Cited by 16 publications

(8 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the future, we will continue to study streamflow forecasting models. For instance, we could apply dynamic selection approaches 44 , 45 to improve the ensemble’s performance in streamflow forecasting, and residual series modeling could be used to improve the accuracy of statistical and machine learning models 46 , 47 . It can make streamflow forecasting more and more accurate.…”

Section: Discussionmentioning

confidence: 99%

Ensemble streamflow forecasting based on variational mode decomposition and long short term memory

Sun

Zhang

Wang

et al. 2022

Sci Rep

View full text Add to dashboard Cite

Reliable and accurate streamflow forecasting plays a vital role in the optimal management of water resources. To improve the stability and accuracy of streamflow forecasting, a hybrid decomposition-ensemble model named VMD-LSTM-GBRT, which is sensitive to sampling, noise and long historical changes of streamflow, was established. The variational mode decomposition (VMD) algorithm was first applied to extract features, which were then learned by several long short-term memory (LSTM) networks. Simultaneously, an ensemble tree, a gradient boosting tree for regression (GBRT), was trained to model the relationships between the extracted features and the original streamflow. The outputs of these LSTMs were finally reconstructed by the GBRT model to obtain the forecasting streamflow results. A historical daily streamflow series (from 1/1/1997 to 31/12/2014) for Yangxian station, Han River, China, was investigated by the proposed model. VMD-LSTM-GBRT was compared with respect to three aspects: (1) feature extraction algorithm; ensemble empirical mode decomposition (EEMD) was used. (2) Feature learning techniques; deep neural networks (DNNs) and support vector machines for regression (SVRs) were exploited. (3) Ensemble strategy; the summation strategy was used. The results indicate that the VMD-LSTM-GBRT model overwhelms all other peer models in terms of the root mean square error (RMSE = 36.3692), determination coefficient (R2 = 0.9890), mean absolute error (MAE = 9.5246) and peak percentage threshold statistics (PPTS(5) = 0.0391%). The addressed approach based on the memory of long historical changes with deep feature representations had good stability and high prediction precision.

show abstract

Section: Discussionmentioning

confidence: 99%

Ensemble streamflow forecasting based on variational mode decomposition and long short term memory

Sun

Zhang

Wang

et al. 2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…A comparison of different ANN-based forecasting models would also be interesting. Support vector regression may also improve the accuracy of the proposed methodology as used in many studies [ [127] , [128] , [129] , [130] , [131] ]. With the increase in available data, time series-based forecasting models can be developed.…”

Section: Discussionmentioning

confidence: 99%

Forecasting COVID-19 recovered cases with Artificial Neural Networks to enable designing an effective blood supply chain

Ayyıldız

Erdoğan

Gümüş

2021

Computers in Biology and Medicine

View full text Add to dashboard Cite

“…As a remedy, several researchers have proposed an ensembling addition to this mixture approach [26,[30][31][32][33]. After the latter model is fit on the residuals of the former model, an ensemble algorithm is introduced to the scheme, which combines the forecasts of the two models in a linear or nonlinear fashion.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Diogo M. F. Izidio et al [32] used a similar method for energy consumption forecasting, where a nonlinear model of choice is fit on the residuals of a seasonal ARIMA model, and a separate nonlinear model is used for combination of the forecasts of the two models. João Fausto Lorenzato de Oliveira et al [33] investigated different techniques to combine the forecasts of different linear and nonlinear models using PSO (Particle Swarm Optimization). However, even though this approach has been widely used by various researchers in different fields, the two base models are still independently optimized to the data, hence, the training is independent of the model combination stage [18].…”

Section: Literature Reviewmentioning

confidence: 99%