Please use this identifier to cite or link to this item:
http://148.72.244.84:8080/xmlui/handle/xmlui/14527
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Maryam Luqman Othman, Nihad S. KhalafAljboori | - |
dc.date.accessioned | 2024-08-06T20:17:03Z | - |
dc.date.available | 2024-08-06T20:17:03Z | - |
dc.date.issued | 2024 | - |
dc.identifier.citation | https://ijas.uodiyala.edu.iq/index.php/IJAS/article/view/36/2 | en_US |
dc.identifier.issn | 3006-5828 | - |
dc.identifier.uri | http://148.72.244.84:8080/xmlui/handle/xmlui/14527 | - |
dc.description.abstract | This paper presents the mathematical structure of the STR decomposition model and the Elman neural network, in addition to the structure of the hybrid model combining the two previous models. The stages of analysis and verification of each model are discussed separately, and the paper proposes the use of the STR decomposition model based on the autoregressive equation and moving averages, while the STR-ENN model is a model thatcombines the STR model and the ENN neural network.The studied data series represents the monthly US soybean oil contracts for the period from 1-2-1997 to 1-5-2022, and using the MATLAB-a2022 program, the results obtained from the hybrid algorithm were compared with the STR model and the ENN neural network individually, to find out which The models are better in terms of prediction, with the prediction accuracy criteria MAE and MSE used. The proposed STR-ENN model had the best predictive performance among the rest of the models, as it had an average absolute error valueless than the other two models | en_US |
dc.language.iso | en | en_US |
dc.publisher | University of Diyala – College of Education for Pure Sciences | en_US |
dc.subject | STR-ENN hybrid | en_US |
dc.subject | STR | en_US |
dc.subject | ENN | en_US |
dc.subject | BP Algorithm | en_US |
dc.subject | RNN | en_US |
dc.title | Preview the predictive performance of the STR, ENN, and STR-ENN hybrid models | en_US |
dc.type | Article | en_US |
Appears in Collections: | المجلة العراقية للعلوم التطبيقية / Iraqi Journal for Applied Science |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.